Reply of Chile

INTERNATIONAL COURT OF JUSTICE
DISPUTE OVER THE STATUS AND USE OF THE
WATERS OF THE SILALA
(CHILE v. BOLIVIA)
REPLY OF THE
REPUBLIC OF CHILE
REPLY AND EXPERT REPORTS
VOLUME 1 OF 3
15 FEBRUARY 2019

1
REPLY OF THE REPUBLIC OF CHILE
VOLUME 1
TITLE PAGE Nº
Reply of the Republic of Chile 3
Expert Report: Wheater, H.S. and Peach, D.W., Impacts of
Channelization of the Silala River in Bolivia on the Hydrology of
the Silala River Basin
85
Expert Report: Peach, D.W. and Wheater, H.S., Concerning the
Geology, Hydrogeology and Hydrochemistry of the Silala River
Basin
155
Statements of Independence and Truth of Drs. Howard Wheater
and Denis Peach
225
List of Annexes to the Reply 227
List of Annexes to the Expert Reports 231
Certification 233
2
3
INTERNATIONAL COURT OF JUSTICE
DISPUTE OVER THE STATUS AND USE OF THE
WATERS OF THE SILALA
(CHILE v. BOLIVIA)
REPLY OF THE
REPUBLIC OF CHILE
REPLY AND EXPERT REPORTS
VOLUME 1 OF 3
15 FEBRUARY 2019

5
iii
REPLY OF THE REPUBLIC OF CHILE
LIST OF FIGURES............................................................................................... vi
LIST OF TABLES ................................................................................................ vi
CHAPTER 1 INTRODUCTION ........................................................................... 1
A. The dispute before the Court .......................................................... 1
B. The structure of the Reply .............................................................. 7
CHAPTER 2 BOLIVIA’S CLAIMS TO THE “ARTIFICIALLY
ENHANCED FLOW” OF THE SILALA RIVER HAVE NO
SUPPORT IN INTERNATIONAL LAW AND IGNORE KEY
HISTORICAL FACTS............................................................................... 9
A. The principles reflected in the Convention on the Law of Non-
Navigational Uses of International Watercourses apply to
international watercourses and the totality of their waters ........... 10
1. International law does not recognize the
concept of “artificial” water ................................. 10
2. The principle of equitable and reasonable
utilization is fully compatible with efforts to
optimize international watercourses..................... 16
3. There is no justification whatsoever for
upstream States to demand compensation for
the construction or maintenance of works
unilaterally implemented within their
territory................................................................. 18
4. The case law, State practice and doctrine
referred by Bolivia do not support the
existence of a distinct legal regime for
“artificially-enhanced flow” and are
irrelevant to Bolivia’s case................................... 28
B. The historical background relevant to Bolivia’s counter-claims:
key omissions by Bolivia.............................................................. 33
6
iv
1. Bolivia ignores almost 100 years of joint
Bolivian-Chilean recognition of the Silala as
a river without distinguishing “natural” from
“artificial” flow .................................................... 34
2. The true facts with respect to the 1906 and
1908 concessions and the later (1928)
channelization in Bolivia for sanitary
reasons.................................................................. 37
3. Bolivia’s failure to take account of the
simple fact that the channels were built with
Bolivian authorization.......................................... 41
4. Notwithstanding the termination in 1997 of
the 1908 concession, Bolivia has not
removed the channels and restored the
wetlands................................................................ 42
C. Conclusion: The distinction between “natural” and
“artificially-enhanced” flow with the legal consequences
alleged by Bolivia is untenable under international law and
Bolivia’s second and third Counter-Claims must be dismissed ... 45
CHAPTER 3 BOLIVIA’S CONTENTIONS ON THE ALLEGED
IMPACT OF THE CHANNELIZATION IN BOLIVIA ARE
UNTENABLE AS A MATTER OF FACT............................................. 47
A. Chile and Bolivia largely agree on the nature and functioning
of the Silala River as an international watercourse ...................... 49
1. Chile and Bolivia agree that the Silala River
is a perennial flow that rises at two sets of
springs in Bolivia and flows along the
natural topographic gradient from Bolivia
into Chile.............................................................. 49
2. Chile and Bolivia agree that the 1928
channelization in Bolivia has only a minor
effect on the direct loss of water to
evaporation of no more than 2% of the
current cross-border flow ..................................... 50
3. Chile and Bolivia agree on the complexity of
the groundwater flow systems of the Silala,
having different origins and recharge areas ......... 51
7
v
4. While Chile and Bolivia maintain different
interpretations of the geology and
hydrogeology of the Silala River basin, this
does not affect their common understanding
of the nature of the Silala River as an
international watercourse ..................................... 52
B. Bolivia’s estimation of the impact of the 1928 channelization
in Bolivia on the cross-boundary surface flows (30-40%
“artificial flow”) is untenable and based on a fundamentally
flawed numerical model ............................................................... 53
1. The three scenarios (“Baseline”, “No Canal”
and “Restored Wetlands”) used by Bolivia to
calculate the 30-40% “artificial flow” are
inconsistent with the law of conservation of
mass and cannot lead to a reliable
calculation ............................................................ 53
2. Bolivia’s estimation is based on a
fundamentally flawed numerical model,
resulting in a gross overestimate of the
impact of the wetland channelization on
surface flow rates, by a factor of about 20 ........... 59
3. The DHI Near Field model is built on an
incorrect interpretation of the geology and
hydrogeology........................................................ 65
4. Any reduction of the cross-boundary surface
flow would anyway be compensated by an
increase of cross-boundary groundwater
flow....................................................................... 67
5. The conclusions of the Ramsar Report on
wetland degradation at the Silala are
unwarranted and are contradicted by recent
evidence provided by DHI and other expert
reports................................................................... 68
C. Conclusion: The impact of the 1928 channelization due to
reduced loss to evapotranspiration, is limited to no more than
2% of the current cross-boundary surface flow; any additional
impact argued by Bolivia is grossly exaggerated ......................... 72
SUBMISSIONS ................................................................................................... 75
8
vi
LIST OF FIGURES
Figure 1. Approximate extents of the Silala Near Field (reproduced from DHI
Report. BCM, Vol. 2, p. 271, Figure 3). ............................................................... 54
Figure 2. (a) A typical groundwater head gradient from the near field model
boundary to the wetland; (b) A typical groundwater head gradient from the far
field model boundary to the wetland (Wheater and Peach (2019), p. 23, Figures 3
(a) and (b))............................................................................................................. 63
LIST OF TABLES
Table 1. DHI’s results of its modelling of different scenarios (reproduced from
DHI Report. BCM, Vol. 5, p. 67, Table 1)............................................................ 55
9
1
CHAPTER 1
INTRODUCTION
1.1 This Reply is submitted in accordance with the time-limits fixed by
the Court in its Order of 15 November 2018, directing the submission of a Reply
by the Republic of Chile, limited to the Counter-Claims presented by the
Respondent.
A. The dispute before the Court
1.2 In its Memorial of 3 July 2017, Chile noted that the dispute before
the Court is straightforward and limited in nature.1 Chile seeks a declaration from
the Court to the effect that the Silala River is an international watercourse (as had
been consistently recognized by both Chile and Bolivia for almost a century prior
to September 1999 when Bolivia abruptly changed its position),2 with the rights
and obligations for its riparian States that arise as a corollary.3 Chile decided to
request such declaration following Bolivia’s President Mr. Evo Morales’ public
announcement in March 2016 that Chile was “stealing” Silala waters from
Bolivia and that Bolivia would present a claim before this Court, and subsequent
statements of the Minister of Foreign Affairs of Bolivia that the presentation of
such claim would take at least two years.4
1.3 Following the lodging of Bolivia’s Counter-Memorial (BCM) of
3 September 2018, the dispute has become even more limited. Bolivia
1 Chile’s Memorial (henceforth “CM”), paras. 1.3 and 1.5.
2 CM, para. 1.8.
3 CM, para. 1.2.
4 CM, paras. 1.8-1.9.
10
2
acknowledges (as it had prior to 1999) that the Silala is indeed an international
watercourse that flows along the natural topographic gradient from Bolivia to
Chile, crossing the border in a natural ravine.5 Bolivia also acknowledges that
both riparian States have rights and obligations with respect to equitable and
reasonable utilization of the Silala, prevention of significant harm, cooperation,
timely notification of planned measures which may have a significant adverse
effect, exchange of data and information and, where appropriate, the conduct of
environmental impact assessments.6
1.4 The issue that is now left for determination is a new assertion by
Bolivia of alleged sovereign rights with respect to a portion of the waters of the
Silala which it characterizes as “artificially-flowing Silala waters” (as opposed to
the Silala’s “natural flow”).7 Bolivia asserts that this “artificial flow” is generated
by the channels and drainage systems located in Bolivia’s territory, and that it
contributes 30-40% of the current transboundary surface flow.8 Bolivia also
contends that customary international law on the use of international
watercourses does not apply to what it calls the “artificial” component of the
Silala flow,9 and that the “delivery” of these “artificial” waters to Chile is subject
to future agreement between the two States.10
1.5 These contentions underpin Bolivia’s defence to Chile’s claims but
also the counter-claims that are the subject of this Reply. In particular, Bolivia
claims that it has sovereignty over the artificial flow of Silala waters engineered,
enhanced, or produced in its territory (Counter-Claim b)), and that any delivery
5 Bolivia’s Counter-Memorial (henceforth “BCM”), para. 44.
6 BCM, paras. 16-18.
7 BCM, para. 14.
8 BCM, para. 13.
9 BCM, para. 14.
10 BCM, para. 20.
11
3
from Bolivia to Chile of such artificially-flowing waters, and the conditions and
modalities thereof, including the compensation to be paid for said delivery, are
subject to the conclusion of an agreement with Bolivia (Counter-Claim c)).11
Bolivia’s Counter-Claim a), which concerns Bolivia’s sovereignty over artificial
channels and drainage mechanisms located in its territory,12 is not contested by
Chile, insofar as Bolivia’s exercise of sovereignty complies with its obligations
regarding the Silala as an international watercourse. Because there is no extant
dispute regarding Bolivia’s sovereignty over its territory, the Court lacks
jurisdiction over Counter-Claim a).13 In the alternative, Counter-Claim a) is
moot.
1.6 Counter-Claims b) and c) (and likewise the parallel defence to
Chile’s claims) have no foundation in fact or in law.
1.7 As to the facts, there is a very basic point that the Silala rises on
Bolivian territory and flows downhill into Chile. Even if it were correct that the
works that Bolivia licensed on its territory had a significant impact on surface
water flows (it is not), absent such works, the same water would anyway flow
down into Chile as groundwater.14 Bolivia has no case, and can have no case that,
absent the works, the so-called “artificial flows” would somehow defy gravity
and remain lodged within that part of the Silala system of ground and surface
waters that is located on Bolivia’s territory.
11 BCM, para. 181 b) (henceforth also second Counter-Claim) and c) (henceforth also third
Counter-Claim).
12 BCM, para. 181 a) (henceforth also first Counter-Claim).
13 Alleged Violations of Sovereign Rights and Maritime Spaces in the Caribbean Sea (Nicaragua v.
Colombia), Counter-Claims, Order of 15 November 2017, I.C.J. Reports 2017, p. 289, at p. 311,
paras. 69-70. Chile considers that this discrete issue of jurisdiction – i.e. the absence of jurisdiction
to rule on Counter-Claim a) because there is no dispute between the Parties as required by
Article XXXI of the Pact of Bogota – can be decided by the Court together with the merits.
14 As Bolivia confirms: “Water on the surface and in the subsurface generally flow in a westward
direction.” BCM, para. 47.
12
4
1.8 Indeed, if Bolivia wishes to remove the channels and to restore the
wetlands to their pre-1920s state, this is something that Chile would positively
encourage:
(a) It is recalled that Chile had no involvement whatsoever in the
construction of the channels in Bolivian territory. The works were
carried out in 1928 by a British company, The Antofagasta (Chili)
and Bolivia Railway Company (the “Railway Company” or
“FCAB”), in pursuit of a concession that had been granted by
Bolivia in 1908. The 1908 concession was terminated unilaterally
by Bolivia in 1997. Since then, nothing has prevented Bolivia from
removing or filling up the channels in order to restore the Cajones
and Orientales wetlands to their natural state.
(b) Chile encourages Bolivia to take all measures necessary to
preserve the wetlands in Bolivia, and if this includes removing the
stone-lined channels built in the 1920s, that would meet with no
objection at all from Chile. Of course, any restoration of the
wetlands would have to be undertaken in a manner not to impair
the natural conditions of the Silala water system, i.e. without
contravening Bolivia’s obligations towards Chile as a riparian state
under customary international law and Chile’s right to equitable
and reasonable utilization of the waters of the Silala River.
1.9 Insofar as it is necessary to look further at the facts (it is not),
Chile’s experts confirm that Bolivia’s estimation of 30-40% “artificiallyenhanced
flow” defies common sense and is, at best, grossly exaggerated. These
estimates are wholly based on a hydrological model developed by Bolivia’s
consultant, the Danish Hydraulic Institute (DHI), from very limited data.
13
5
According to Chile’s experts, the modelling has fundamental flaws and uses
unacceptable premises, leading to inaccurate and misleading results.
1.10 As to the law, there is no basis for the distinction made by Bolivia
between “natural flow” and “artificially-enhanced flow”. The principles of the
Convention on the Law of Non-Navigational Uses of International Watercourses
(“UNWC” or “Convention”) apply to international watercourses and the totality
of their waters without distinction. The impracticality of Bolivia’s theory is
underscored by the fact that Bolivia at no point indicates how to separate the
“natural” from the “artificial” flows in the Silala River, nor how it purports to
“deliver” the “artificial flow” under a supposed future agreement, in
circumstances where all the water anyway flows inevitably into Chile due to the
topographical gradient. Any increase in surface water flow due to the
channelization would result in an almost equivalent decrease in groundwater
flow, and the overall cross-boundary flow into Chile would remain practically the
same.15
1.11 Moreover, the optimization of an international watercourse by
upstream States, as for instance by canal lining or more efficient upstream uses,
does not set aside the fundamental principle of equitable and reasonable
utilization of shared watercourses or give rise to a right to compensation. If such
were the case, upstream States could impose a “water tax” on downstream States
by optimizing their water usage and letting more water pass through, which is not
acceptable under customary international law.
15 Additional loss to evaporation in the no-channel scenario will be no more than 2% of the flow as
agreed by both Bolivia’s and Chile’s experts (Chile’s recent evidence suggests that changes are
non-existent; Bolivia’s wetlands have higher evaporation than an undisturbed wetland in Chile).
Other losses of surface water flow due to channelization effects on groundwater recharge have
similarly been shown to be small and would in any case flow to Chile as groundwater. See
Wheater, H.S. and Peach D.W., Impacts of Channelization of the Silala River in Bolivia on the
Hydrology of the Silala River Basin (henceforth “Wheater and Peach (2019)”), pp. 4-5.
14
6
1.12 Take as a hypothetical example, State A, which carries out certain
channelization works on the stretch of an international watercourse in its territory,
with the result that 20% less water dissipates in its territory and instead flows
downstream into the territory of State B. State A does not thereby create
sovereign rights and/or some right to compensation with respect to that water.
Any contention to the contrary betrays a fundamental misunderstanding of the
law relating to international watercourses.
1.13 The inappropriateness of such a scheme is even more evident in the
present case, where the construction of the channels that allegedly increased the
water flow (although Chile’s experts maintain that the effect is negligible) was
carried out by a private company and authorized by Bolivia pursuant to a
Bolivian concession, without any prior consultation with Chile.
1.14 Chile wishes to reassure Bolivia that it fully recognises Bolivia’s
sovereignty over the artificial channels and drainage mechanisms in the Silala
that are located in its territory, and the right to decide whether and how to
maintain them (Counter-Claim a)). Again, Chile encourages Bolivia to restore the
wetlands, as it appears in its formulation of Counter-Claim a) that Bolivia wishes
to do, in so far as this complies with Bolivia’s obligations towards Chile under
customary international law.
1.15 In addition, Chile wishes to reassure Bolivia that it does not claim
to pre-empt any future uses by Bolivia of the Silala River, to the extent that such
uses are consistent with the principle of equitable and reasonable utilization, and
provided that Bolivia complies with its obligation under customary international
law to prevent the causing of significant harm and related obligations concerning
cooperation, notification, exchange of information and, where appropriate, the
conduct of environmental impact assessment, in accordance with customary
international law. This is important, because Bolivia questions whether Article 11
15
7
of the UNWC is part of customary international law and also denies that these
related obligations have been engaged in the present circumstances.
1.16 In light of all the above, the dispute before the Court has been very
significantly reduced as compared to when Chile decided to lodge its Application
in June 2016. Bolivia now recognizes that the Silala River is an international
watercourse. Bolivia’s defence and Counter-Claims b) and c), building on the
notion of an “artificially-enhanced watercourse”, are legally and factually
untenable. They should be dismissed by the Court.
B. The structure of the Reply
1.17 The structure of this Memorial is as follows: chapter 2 explains in
greater detail the untenable nature of Bolivia’s thesis that international law
distinguishes between an international watercourse and an “artificially enhanced
watercourse”. Chile also points out that any “artificially enhanced flow” in the
Silala River is attributable to the acts of Bolivia. In chapter 3 Chile addresses the
limited differences between the Parties as to the facts, noting however that these
are not dispositive of the case (which is dealt with entirely by chapter 2). It is
demonstrated that the percentage of “artificially-enhanced flow”, if such flow
exists at all, is grossly overstated and the result of a fundamentally flawed
hydrological model.
1.18 This Reply is supported by two expert reports by Drs. Howard
Wheater and Denis Peach that point out the fundamental flaws in the hydrological
model developed by Bolivia’s consultant DHI. They also provide additional data
to support and/or refine the conclusions reached in their earlier expert reports
submitted together with Chile’s Memorial (CM) of 3 July 2017. The Wheater and
16
8
Peach reports are in turn supported by a number of underlying studies into the
Silala River system that are annexed to the Reply.
17
9
CHAPTER 2
BOLIVIA’S CLAIMS TO THE “ARTIFICIALLY ENHANCED FLOW”
OF THE SILALA RIVER HAVE NO SUPPORT IN INTERNATIONAL
LAW AND IGNORE KEY HISTORICAL FACTS
2.1 In its Counter-Memorial, Bolivia reverts to its pre-1999 recognition
that the Silala is an international watercourse and that, as such, its use is governed
by the rules of international law concerning international watercourses. This
acceptance of what is obvious has left Bolivia in a difficulty, given that it has
elected to pursue its defence and make a counter-claim: how to assert control and
sovereign rights over water that Bolivia is not using, and which international law
requires it to share in an equitable and reasonable manner with Chile (which
alone is using the water).
2.2 In an effort to escape this difficulty, Bolivia invents a notion
unfounded in science or law, namely, that works that Bolivia had authorized in
Bolivian territory – chiefly the excavation of earth channels in the wetlands, some
of which were lined with stone, and lining of the Silala natural river channel –
produced an “artificial flow” over which “Bolivia has sovereignty” and for which
Chile must pay compensation.16 According to Bolivia, the “delivery” of this
“artificial flow” from Bolivia to Chile and the conditions and modalities thereof,
“including the compensation to be paid” therefor, “are subject to the conclusion
of an agreement with Bolivia.”17
2.3 As is discussed in section A below, Bolivia’s thesis finds no
support in international law. In section B, Chile points out key omissions in
16 BCM, para. 181 b).
17 BCM, para. 181 c).
18
10
Bolivia’s account of the historical background relevant to its counter-claims that
further undermine its case.
A. The principles reflected in the Convention on the Law of Non-
Navigational Uses of International Watercourses apply to international
watercourses and the totality of their waters
2.4 Chile will show that there is no basis in international law for
distinguishing between natural and “artificial” flows (section 1) and that any such
distinction runs counter to the principle of equitable and reasonable utilization of
shared watercourses (section 2). Chile will also establish that Bolivia’s third
Counter-Claim for compensation for the upkeep of unilaterally instituted
waterworks in its own territory would be not only unprecedented but also
seriously disruptive of existing water governance regimes (section 3). Finally,
Chile will show that the case law, State practice and doctrine referred to by
Bolivia do not support the existence of a distinct legal regime for “artificiallyenhanced”
flow (section 4).
1. International law does not recognize the concept of “artificial” water
2.5 Bolivia takes great pains to try to establish that there are two
separate flows of water in the Silala that cross the border into Chile: a “natural”
flow and an “artificial” or “artificially enhanced” flow.18 It contends that, while
international law governs the “natural” flow, Bolivia “has sovereignty over the
artificial flow” and any “delivery” of this “artificial” flow to Chile is subject to
the conclusion of an agreement between the two countries.19
18 BCM, Chapter 2, in particular section C, “Artificial Enhancement of the Silala”.
19 BCM, para. 181.
19
11
2.6 This argument amounts to nothing less than a denial of the fact that
water flows downhill. All of the natural recharge from precipitation over the
Silala groundwater catchment area will cross the international border, either as
surface water or groundwater.20 The so-called “intake mechanism” and “complex
system of artificial channels and drainage mechanisms within Bolivian territory
near the bofedales”21 that according to Bolivia “produced” the “artificiallyenhanced”
flow, were constructed by the British private Railway Company
FCAB, with the authorization of Bolivia.22 These small channels (about 0.6 m
depth and 0.6 m width)23 could only have very minor impacts on the flow of the
Silala water, which was and always will be down a gradient toward Chile.
2.7 Development of watercourses often takes the form of what the
UNWC refers to as “regulation”.24 Many intensively used watercourses are
regulated in some way or another, often by straightening their channels to
eliminate naturally-occurring meanders and thus facilitate their use for such
purposes as navigation and hydroelectric power production.25 This does not
convert the water they carry from being “natural” to being “artificial.” Even the
bypass canal involved in the Gabčíkovo-Nagymaros Project case,26 which was
constructed by Czechoslovakia, runs for 31 km through what is now Slovak
20 Wheater and Peach (2019), p. 4.
21 BCM, para. 63.
22 Deed of Concession by the State of Bolivia of the Waters of the Siloli (No. 48) to The
Antofagasta (Chili) and Bolivia Railway Company Limited, 28 October 1908. CM, Vol. 3,
Annex 41.
23 Wheater, H.S. and Peach, D.W., The Silala River Today – Functioning of the Fluvial System
(Exp. Rep. 1), p. 6. CM, Vol. 1, p. 134.
24 Convention on the Law of the Non-Navigational Uses of International Watercourses (henceforth
“UNWC”), signed at New York on 21 May 1997, U.N. Doc. A/RES/51/229 (1997), Art. 25. CM,
Vol. 2, Annex 5.
25 This is the case, for example, with the Danube. See Gabčíkovo-Nagymaros Project
(Hungary/Slovakia), Judgment, I.C.J. Reports 1997, p. 7 (henceforth “Gabčíkovo-Nagymaros
Project”).
26 Ibid.
20
12
territory, and is capable of carrying some 80 to 90% of the Danube’s flow, was
never thought to carry “artificial” flows. Nor of course could any claim have been
made by Slovakia for compensation from Hungary for any “artificially enhanced”
flows carried by the bypass canal following Hungary’s attempted termination of
the 1977 treaty involved in the case.
2.8 In addition, domestic courts in disputes involving similar factual
situations have not found that a user who has altered a watercourse system to
reduce water “loss” to evaporation and evapotranspiration gains a water right in
the net gain to the stream of the kind claimed by Bolivia. The facts of the United
States case R.J.A., Inc. v. The Water Users Association of District No. 6, et al.,27
are similar to the alleged facts of the present dispute. The case involved property
situated in the U.S. State of Colorado at the headwaters of a tributary of the South
Platte River. The plaintiff company undertook a “project that will reduce water
loss from a marshy mountain meadow by removing the underlying peat moss,
thereby eliminating a saturated, seepy condition. This will decrease evaporation
from the soil and surface and reduce evapotranspiration from grassy
vegetation.”28 The plaintiff asserted a right to the water thus saved – as does
Bolivia to the “artificially-enhanced” flows as a result of its channelization and
stone lining works.29
27 R.J.A., Inc. v. The Water Users Association of District No. 6, et al., Supreme Court of Colorado,
Sep. 10, 1984, 690, P.2d 823 (1984). Available at: https://casetext.com/case/rja-inc-v-water-usersassoc.
28 Ibid., p. 824. Specifically, the plaintiff’s property “originally included a 27-acre [10.9 hectare]
peat moss marsh which was approximately 3000 years old and, thus, was in existence long before
any water rights were established on the [relevant] River system. [...] According to the [plaintiff],
loss of water to the atmosphere was higher from this peat moss marsh than from a well-drained
mountain meadow of equivalent size. [...] In the early 1970s, the [plaintiff] undertook a project to
remove the extensive deposits of peat moss underlying the marsh, drain the land, and convert the
marsh to a well-drained meadow [...].”
29 Ibid. Plaintiff claimed that “the drainage of the marsh and elimination of the saturated, seepy
condition would reduce the rates of evaporation and evapotranspiration, and thereby would
decrease consumptive use of water by 43.3 acre feet [53,409,764.33 liters] per year. Because this
21
13
2.9 The Supreme Court of Colorado, in an en banc decision,30 ruled
against the plaintiff, holding that “reduction of consumptive use of tributary water
cannot provide the basis for a water right that is independent of the system [of
water rights] on the stream.”31 The court referred to its jurisprudence
distinguishing between “developed” and “salvaged” water. “Developed” water is
“new water not previously part of the river system, i.e., it is imported or nontributary
water.”32 “Salvaged” water is “tributary water made available for
beneficial use through elimination of waste.” The court explained that “[o]nly
developed water can be made the basis of a right independent of the [otherwise
existing water rights] system,”33 and made plain that it was not willing to “create
a superclass of water rights never before in existence”.34
2.10 As for this case, to hold –as Bolivia pretends– that the construction
of works to optimize the quality of the water, which might also have had a minor
effect on the optimization of the flow, would likewise be to “create a superclass
of water rights never before in existence” in international law.35 Such a
“superclass” of rights would disrupt established rights and obligations in
would represent a net gain to the stream, the [plaintiff] asserted that its water right should not be
subject to administration under the [applicable water rights] system.”
30 Ibid. An “en banc” decision is one by all of the judges of a court rather than the more common
case of a hearing and decision by a panel of the court. Cases heard en banc are often ones of
exceptional public importance.
31 Ibid., p. 825.
32 Ibid.
33 Ibid.
34 Ibid., p. 827, quoting from Southeastern Colorado Water Conservancy District v. Shelton
Farms, Inc., 187 Colo. 181, 190, 529 P.2d 1321, 1326 (1975), the court noted that it had there
“concluded that, since the water in question had always been tributary to the stream and was not
water new to the river system, the developed water cases were inapposite. [...] ‘To hold any other
way would be to weaken the [water rights] system, and create a superclass of water rights never
before in existence.’”
35 Ibid.
22
14
international watercourses throughout the world and is clearly unacceptable under
customary international law.
2.11 Finally, Bolivia appears to be concerned that the channelization of
the Silala might qualify as a “canal” and as such be included in the International
Law Commission’s commentary to its 1994 draft articles on international
watercourses, as one of the possible components of a hydrologic system
constituting a “watercourse.”36 Thus, Bolivia tries to make good the argument
that the International Law Commission (ILC) did not intend to include “artificial
diversions” in its definition of a “watercourse”.37
2.12 Bolivia’s contention that the ILC’s definition of watercourse does
not include “artificial diversions” is misplaced, for several reasons. First, the oral
exchange to which Bolivia refers in support of this contention, published in the
summary records of the ILC’s 1987 session,38 occurred while the draft articles
were still in gestation and does not reflect any concluded position. Second, the
exchange related to the definition of the expression “watercourse States,” not the
term “watercourse.” Indeed, the ILC adopted several introductory articles at that
session, which were later revised in its draft articles adopted on second reading in
1994, but which did not include a definition of the term “watercourse.” The
Commission postponed defining that term until work on the entire set of draft
articles was otherwise complete. Third, the “personal” view of the special
rapporteur referred to by Bolivia39 was just that, and not necessarily the view of
the Commission.40
36 Yearbook of the International Law Commission, 1994, vol. II (Part Two), p. 90, para. 4 of the
commentary.
37 BCM, para. 96.
38 Yearbook of the International Law Commission, 1987, vol. I, p. 220, para. 75.
39 BCM, para. 96. The opinion in question was that the special rapporteur would be “reluctant to
define international watercourse so as to include such man-made diversions as a canal, which
23
15
2.13 In seeking to establish that canals are not part of an international
watercourse and not subject to the rules of customary international law,41 Bolivia
also quotes selectively from the authorities on which it relies. The quotation from
the Max Planck Encyclopedia of Public International Law,42 for example,
neglects to include the part of the entry concerning canals and non-navigational
uses, the material most applicable to the present case.43
2.14 In any event, the waterworks in Bolivia, consisting of the
excavation of earth channels in the wetlands and straightening and lining of the
natural river channel, do not come close to qualifying as a “canal” as understood
in international practice.44 Thus, Chile does not contend that the water channels
might take the water of an international watercourse into another drainage basin.” Yearbook of the
International Law Commission, 1987, vol. I, p. 220, para. 75. This provisional view stated orally
during the Commission’s debate in 1987, in answer to a question, related to a canal used for a
specific purpose.
40 The special rapporteur’s personal view was supplanted by the commentaries the ILC adopted in
both 1991, on first reading, and in 1994, on second reading, which included “canals” as one of the
possible components of a hydrologic system, see Yearbook of the International Law Commission,
1994, vol. II (Part Two), p. 90, para. 4 of the Commentary.
41 BCM, paras. 94-102.
42 BCM, para. 99.
43 The relevant passage reads: “When a canal is so constructed as to affect an international
watercourse and its water resources, it may be subject to the rules of customary international law
governing the non-navigational uses of international watercourses. In the context of the
codification of that body of law [...], the International Law Commission (ILC) has defined an
international watercourse as a system of surface waters and groundwaters constituting by virtue of
their physical relationship a unitary whole, the parts of which are situated in different States, and
whose components can include rivers, lakes, aquifers, glaciers, reservoirs and canals. [...] As this
definition is also embodied in Art. 2 Convention on the Law of the Non-Navigational Uses of
International Watercourses, it follows that the general principles codified therein, and in particular
the rule of equitable utilization of shared water resources and the obligation not to cause
significant harm to other riparian States, can apply to canals integrated to an international
watercourse system and exploited for non-navigational purposes.” (Emphasis added) M. Arcari,
“Canals”, Max Planck Encyclopaedia of Public International Law, online version, last updated
October 2007, para. 9. Available at: http://opil.ouplaw.com/view/10.1093/law:epil/
9780199231690/law-9780199231690-e1013?rskey=GDqdM0&result=1&prd=EPIL.
44 In international practice the term “canal” is normally used to refer to a means of water
conveyance that is separate from the bed of a river. The Rhine-Main-Danube canal, which was
completed in 1992 and links two major international drainage basins, those of the Rhine and the
Danube, is an example. It would be considered to be part of an international watercourse although
24
16
in Bolivia constitute one or more “canals”. However, the ILC’s inclusion of
canals as a possible component of a watercourse means that, a fortiori, the stone
lining of a stream would not in any way impact upon its character as a
“watercourse.” It is therefore plainly wrong that, as Bolivia contends, “the
evidence indicates that the accepted norm is to exclude artificial conveyance
mechanisms like canals and drainage mechanisms from the scope of customary
international law applicable to transboundary watercourses.”45
2. The principle of equitable and reasonable utilization is fully compatible
with efforts to optimize international watercourses
2.15 Bolivia’s argument that optimization of the flow of the Silala
creates “artificially-enhanced” water flows that are not governed by the
customary international law principle of equitable and reasonable utilization runs
counter both to that principle and to State practice.
2.16 Article 5 of the UNWC, after setting forth the principle of equitable
and reasonable utilization, states as follows:
“In particular, an international watercourse shall be used and developed
by watercourse States with a view to attaining optimal and sustainable
utilization thereof and benefits therefrom, taking into account the
interests of the watercourse States concerned, consistent with adequate
protection of the watercourse.”46 (Emphasis added)
2.17 The ILC’s commentary to Article 5 explains that the attainment of
“optimal utilization” of an international watercourse “implies attaining maximum
possible benefits for all watercourse States and achieving the greatest possible
it is called a “canal.” Information available at: https://www.britannica.com/topic/Main-Danube-
Canal.
45 BCM, para. 101.
46 UNWC, Article 5(1). CM, Vol. 2, Annex 5.
25
17
satisfaction of all their needs, while minimizing the detriment to, or unmet needs
of, each.”47 This is to be done in a manner that is “consistent with adequate
protection of the watercourse.”48
2.18 As the ILC’s commentary to Article 25 states, “Regulation of the
flow of watercourses is often necessary […] to maximize the benefits that may be
obtained from the watercourse.”49 As stated two decades earlier by the Indian
Krishna Water Disputes Tribunal: “Needless waste of water should be prevented
and efficient utilisation encouraged”.50
2.19 Article 25(3) of the Convention defines “regulation” to mean “the
use of hydraulic works or any other continuing measure to alter, vary or
otherwise control the flow of the waters of an international watercourse.”51 This
definition readily encompasses the minor hydraulic works installed by the
Railway Company in Bolivia that were designed to preserve the quality of the
Silala River waters.52 The fact that this form of development of a watercourse is
recognized in the Convention as a form of use by States that is governed by
international law is in itself a sufficient answer to Bolivia’s claim that the
regulatory works on the Silala in its territory are “artificial conveyance
mechanisms like canals and drainage mechanisms [that are excluded] from the
47 Yearbook of the International Law Commission, 1994, vol. II (Part Two), p. 97, para. (3) of
commentary to Article 5.
48 UNWC, Article 5(1). CM, Vol. 2, Annex 5.
49 Yearbook of the International Law Commission, 1994, vol. II (Part Two), p. 126, para. (1) of
commentary to Article 25.
50 Krishna Water Disputes Tribunal, Decision of 24 December 1973, para. 310. Available at:
http://cwc.gov.in/main/downloads/KWDT%201volume1.pdf.
51 UNWC, Article 25(3). CM, Vol. 2, Annex 5.
52 As Chile has demonstrated, the channelization was carried out for sanitary reasons, not to
increase the flows, see CM para. 4.61. Moreover, the channelization counted with Bolivia’s
authorization under the 1908 Bolivian concession, see CM, Vol. 3, Annex 41.
26
18
scope of customary international law applicable to transboundary
watercourses.”53
2.20 Even if it were the case, as Bolivia suggests, that “[t]he
channelization system was installed to improve the transport of Silala water into
Chile […] necessary to create a more consistent and voluminous flow of water
from the Silala springs in Bolivia, through the dense bofedales, and across the
border into Chile”,54 these purposes are covered by the concept of regulation, an
activity recognized by the UNWC as being governed by international law and
consistent with the principle of equitable and reasonable utilization. Hence, the
works by the Railway Company in Bolivia, designed as they were to preserve the
quality of the Silala waters, are in conformity with the requirements of Article 5
of the Convention.
3. There is no justification whatsoever for upstream States to demand
compensation for the construction or maintenance of works unilaterally
implemented within their territory
2.21 In its third Counter-Claim, Bolivia states:
“c) Any delivery from Bolivia to Chile of artificially-flowing waters of
the Silala, and the conditions and modalities thereof, including the
compensation to be paid for said delivery, are subject to the conclusion
of an agreement with Bolivia.”55
2.22 It has been shown in section 1 above that international law does not
recognize the concept of “artificially flowing waters” and, in section 2, that there
can be no legal basis for Bolivia’s contention that it has “sovereignty” over an
53 BCM, para. 101.
54 BCM, para. 53.
55 BCM, para. 181 c).
27
19
alleged “artificial” flow.56 Nonetheless, Bolivia goes on to demand in its third
Counter-Claim that Chile conclude an agreement as a precondition to the
“delivery” by Bolivia of the “artificially-flowing waters of the Silala”.57 This
section will address Bolivia’s third Counter-Claim.
a. Bolivia’s third Counter-Claim in effect asserts a right of veto over
Chile’s right to receive Silala waters
2.23 Through this Counter-Claim, Bolivia asserts that it has the right to
effect the “delivery” of the so-called “artificial flow” of Silala waters to Chile
pursuant to terms that are ultimately subject to Bolivia’s agreement. Included as
one of these terms is “the compensation to be paid [by Chile to Bolivia] for said
delivery.”58
2.24 This assertion has no basis in reason or law. In the well-known
Lake Lanoux Arbitration,59 Spain asserted that France’s right to develop the
Carol River, which flows from France into Spain, was subject to the prior
agreement, or consent, of Spain. The arbitral tribunal disagreed. It stated:
“[T]o evaluate in its essence the need for a [prior] agreement, it is
necessary to adopt the hypothesis that the States concerned cannot
arrive at an agreement. In that case, it would have to be admitted that a
State which ordinarily is competent has lost the right to act alone as a
consequence of the unconditional and discretionary opposition of
another State. This is to admit a ‘right of consent’, a ‘right of veto’,
56 BCM, paras. 180 and 181 b).
57 BCM, para. 181 c). See also Submission 2 c), BCM, p. 106.
58 BCM, para. 165.
59 Affaire du Lac Lanoux (Spain v. France), Award of 16 November 1957, Reports of International
Arbitration Awards, Vol. XII, p. 281. English translations in 24 ILR p. 101 (1961); 53 AJIL p. 156
(1959); and Yearbook of the International Law Commission, 1974, vol. II (Part Two), p. 194.
28
20
which at the discretion of one State paralyses another State’s exercise of
its territorial competence.”60
2.25 In the present case it is the upstream State, Bolivia, that is claiming
a right of consent to the “delivery” into Chile of what Bolivia calls the
“artificially-enhanced” flow of the Silala. But the effect is the same: Bolivia is
claiming “a ‘right of veto’, which at [Bolivia’s] discretion […] paralyses
[Chile’s] exercise of its territorial competence” in accordance with international
law. Such a claim is unsustainable for the reasons expressed in the Lake Lanoux
award. If the claim were upheld, it would free upstream States to assert a right of
prior consent to downstream States’ use of a shared watercourse and demand
compensation for the release of water into downstream States on the pretext that
the upstream State had created an “artificially-enhanced” flow of that water. This
would be contrary to established principle and would destabilize water relations
around the world.
b. The concept of territorial sovereignty is inapplicable to a shared
natural resource
2.26 Bolivia’s claims regarding Silala flows passing through the works
in its territory are premised on Bolivia’s assertion that it “has sovereignty over
the artificial flow of Silala waters […] and Chile has no right to that artificial
flow; […].”61
2.27 The notion that a State can have exclusive sovereignty over
something that it shares with another State, here freshwater resources, is not
supported by State practice, including the sources cited by Bolivia, as will be
shown in section 4 below. Chile does not dispute Bolivia’s Counter-Claim a) that
it “has sovereignty over the artificial channels and drainage mechanisms in the
60 Yearbook of the International Law Commission, 1994, vol. II (Part Two), p. 197, para. 1065.
61 BCM, para. 181 b).
29
21
Silala that are located in its territory and therefore has the right to decide whether
and how to maintain them.”62 This is in essence a claim that Bolivia has
sovereignty over its territory. Of course Bolivia does, but this is without prejudice
to Chile’s rights as downstream riparian to the equitable and reasonable use of the
waters of the Silala, the prevention of significant harm principle, and Bolivia’s
procedural obligations under customary international law. It follows that Chile
does not contest Counter-Claim a), although there are two important points to be
made.
2.28 First, there is a major difference, and a legally dispositive one,
between the physical works in Bolivia on the one hand, and the waters of the
Silala, on the other. It is impossible for a State to have exclusive sovereignty over
a resource that is shared with another State without doing violence to the concept
of sovereignty.
2.29 That a State does not have exclusive sovereignty over the portion
of an international watercourse within its borders is the basis of the law of
international watercourses, by which Bolivia accepts that it is bound in relation to
the Silala.63 The most fundamental principle in that field of law, equitable and
reasonable utilization,64 is the negation of the notion of exclusive sovereignty
over a portion of an international watercourse within a State’s territory. As
established by the Krishna Water Disputes Tribunal: “No State has a proprietary
interest in a particular volume of water of an inter-State river on the basis of its
contribution [...].”65 Even if, quod non, any portion of the flow of the Silala is
62 BCM, para. 181 a).
63 BCM, paras. 14-16.
64 The Court referred to this principle a conferring a “basic right” on Hungary at co-riparian with
Slovakia of the Danube. Gabčíkovo-Nagymaros Project, at p. 54, para. 78.
65 Krishna Water Disputes Tribunal, Decision of 24 December 1973, para. 308.
30
22
“enhanced” as Bolivia contends, that is still part of the Silala system of waters, a
system that Bolivia shares with Chile.66
2.30 The special meaning of sovereignty in the law of international
watercourses is informed by the Permanent Court’s judgment in the River Oder
case.67 In addressing whether the principle of freedom of navigation provided
downstream States with access to portions of tributaries situated wholly in
upstream States, that Court emphasized the existence of a “community of interest
of riparian States”, rather than fall back on the notion of sovereignty:
“when consideration is given to the manner in which States have
regarded the concrete situations arising out of the fact that a single
waterway traverses or separates the territory of more than one State, and
the possibility of fulfilling the requirements of justice and the
considerations of utility which this fact places in relief, it is at once seen
that a solution of the problem has been sought not in the idea of a right
of passage in favour of upstream States, but in that of a community of
interest of riparian States. This community of interest in a navigable
river becomes the basis of a common legal right, the essential features
of which are the perfect equality of all riparian States in the user of the
whole course of the river and the exclusion of any preferential privilege
of any one riparian State in relation to the others.”68
2.31 In the Gabčíkovo-Nagymaros Project case the Court, after quoting
this passage, stated:
“Modern development of international law has strengthened this
principle for non-navigational uses of international watercourses as
well, as evidenced by the adoption of the Convention of 21 May 1997
66 CM, paras. 2.3-2.6.
67 Case relating to the Territorial Jurisdiction of the International Commission of the River Oder
(Czechoslovakia, Denmark, France, Germany, Great Britain, and Sweden/Poland), 1929, P.C.I.J.,
(Ser. A) No. 23 (Sept. 10), p. 5.
68 Ibid., p. 27.
31
23
on the Law of the Non-Navigational Uses of International Watercourses
by the United Nations General Assembly.”69
2.32 There is thus a community of interest among the riparian States,
Bolivia and Chile, in the shared freshwater resources of the Silala River. Bolivia
must therefore respect Chile’s right to the reasonable and equitable use of Silala
waters – all Silala waters, including any that may allegedly have been “saved” by
the works constructed in Bolivia.
2.33 Second, and as part of the above, Bolivia is obliged to provide
timely notification and consult Chile with respect of planned measures on the
Silala watercourse that may be adversely affected by them.70 Bolivia has
indicated that it accepts this obligation in general terms,71 although its
understanding of what is required does not accord with that of Chile.
2.34 It is not the function of this Reply to respond to Bolivia’s
inadequate defence to Chile’s claims concerning Bolivia’s failure to comply with
its obligations of notification and consultation with respect to a number of past
projects.72 However, in response to Bolivia’s Counter-Claims b) and c), Chile
notes that it is not tenable to contend that the procedural obligations to notify and
69 Gabčíkovo-Nagymaros Project, at p. 56, para. 85. The concept of “community of interest” has
also been invoked by States in disputes before the Court. See, e.g., the Court’s 2006 Provisional
Measures Order in the Pulp Mills case, p. 122, para. 39, and p. 130, para. 64. Pulp Mills on the
River Uruguay (Argentina v. Uruguay), Provisional Measures, Order of 13 July 2006, I.C.J.
Reports 2006, p. 113.
70 Part III of the UNWC, “Planned Measures,” sets forth procedures aimed at preventing
transboundary harm, “assist[ing] watercourse States in maintaining an equitable balance between
their respective uses of an international watercourse” and thus “help[ing] to avoid disputes relating
to new uses of watercourses.” Yearbook of the International Law Commission, 1994, vol. II
(Part Two), p. 111, para. (1) of commentary to Article 12.
71 BCM, para. 153.
72 Three projects were announced by the Governor of the Department of Potosí in 2011, namely,
the construction of a fish farm, a small dam and a mineral water bottling plant. See Note
N° 199/39 from the General Consulate of Chile in La Paz to the Ministry of Foreign Affairs of
Bolivia, 7 May 2012. CM, Vol. 2, Annex 34. The Bolivian Military Post was constructed in 2006
and ten houses near the Military Post were constructed in 2016.
32
24
consult are only applicable to the naturally-flowing Silala waters and not to the
“artificially-enhanced” flows of the Silala,73 and likewise that Article 11 of the
UNWC, “Information concerning planned measures,” does not reflect customary
international law and is not applicable in the present case.74 Article 11 is the
chapeau to Part Three of the Convention, Planned Measures, which is
“introduce[d]”75 by Article 12, “Notification concerning planned measures with
possible adverse effects,” that Bolivia does accept as customary international law.
It follows that Bolivia’s position is also internally inconsistent.
2.35 Bolivia’s sovereignty as asserted in its Counter-Claim a) is of
course subject to the procedural obligations it has accepted, as well as the suite of
related obligations recognized by the Court. Bolivia acknowledges the Court’s
finding that “the obligation to notify is ‘an essential part of the process leading
the parties to consult in order to assess the risks of the plan and to negotiate
possible changes which may eliminate those risks or minimize their effects.’”76
With respect to any work Bolivia may undertake on the Silala to “modify the
artificial channels and drainage mechanisms which are located in its territory in
order to fulfil [the] goal” of maintaining the natural ecology,77 it would be
incumbent upon Bolivia to follow this finding by the Court. Bolivia would
equally be bound to follow the process outlined by the Court in the Certain
Activities and Road cases in order to comply with its obligation of harmprevention:
73 BCM, para. 148.
74 BCM, paras. 153-155.
75 Yearbook of the International Law Commission, 1994, vol. II (Part Two), p. 111, para. 1 of
commentary to Article 12.
76 BCM, para. 156, quoting Pulp Mills on the River Uruguay (Argentina v. Uruguay) Judgment,
I.C.J. Reports 2010, p. 14 (henceforth “Pulp Mills”), at p. 59, para. 115.
77 BCM, para. 180.
33
25
“[T]o fulfil its obligation to exercise due diligence in preventing
significant transboundary environmental harm, a State must, before
embarking on an activity having the potential adversely to affect the
environment of another State, ascertain if there is a risk of significant
transboundary harm, which would trigger the requirement to carry out
an environmental impact assessment.”78
2.36 The Court had found in Pulp Mills case that “it may now be
considered a requirement under general international law to undertake an
environmental impact assessment where there is a risk that the proposed
industrial activity may have a significant adverse impact in a transboundary
context, in particular, on a shared resource.”79 Thus Bolivia would be under an
obligation to prepare an environmental impact assessment in respect of any work
on the Silala River that meets these conditions. Article 18(1) of the UNWC
establishes a process to be followed if Bolivia fails to observe these obligations:
“1. If a watercourse State has reasonable grounds to believe that another
watercourse State is planning measures that may have a significant
adverse effect upon it, the former Sate may request the latter to apply
the provisions of article 12. The request shall be accompanied by a
documented explanation setting forth its grounds.”80
78 Certain Activities Carried Out by Nicaragua in the Border Area (Costa Rica v. Nicaragua) and
Construction of a Road in Costa Rica along the San Juan River (Nicaragua v. Costa Rica),
Judgment, I.C.J. Reports 2015, p. 665, at p. 706, para. 104.
79 Pulp Mills, at p. 83, para. 204. The Court found in its judgment in Certain Activities and
Construction of a Road cases that “Although the Court’s statement in the Pulp Mills case refers to
industrial activities, the underlying principle applies generally to proposed activities which may
have a significant adverse impact in a transboundary context.” Certain Activities Carried Out by
Nicaragua in the Border Area (Costa Rica v. Nicaragua) and Construction of a Road in Costa
Rica along the San Juan River (Nicaragua v. Costa Rica), Judgment, I.C.J. Reports 2015, p. 665,
at p. 706, para. 104.
79 Pulp Mills, at p. 83, para. 204.
80 UNWC, Article 18, para. 1. CM, Vol. 2, Annex 5.
34
26
2.37 If, as has happened in the past,81 Bolivia responds to Chile’s
request by refusing to provide any information on the projects, the remainder of
Article 18 sets forth the procedure to be followed:
“2. In the event that the State planning the measures nevertheless finds
that it is not under an obligation to provide a notification under article
12, it shall so inform the other State, providing a documented
explanation setting forth the reasons for such finding. If this finding
does not satisfy the other State, the two States shall, at the request of
that other State, promptly enter into consultations and negotiations in
the manner indicated in paragraphs 1 and 2 of article 17.82
3. During the course of the consultations and negotiations, the State
planning the measures shall, if so requested by the other State at the
time it requests the initiation of consultations and negotiations, refrain
from implementing or permitting the implementation of those measures
for a period of six months unless otherwise agreed.”83
2.38 This process has not eventuated in respect of past projects, due to
Bolivia’s refusal to provide information concerning them. With regard to any
81 Bolivia claims to have responded to Chile’s repeated requests for information by Notes of
7 May 2012 and 9 October 2012. See CM, Vol. 2, Annexes 34 and 35, and also BCM, paras. 143-
147. However, Bolivia’s Note of 24 May 2012, cited by Bolivia in this context, did not provide
any information on the projects that had been announced by the Governor of Potosí, and on which
Chile had requested information. Rather, it insisted that the Silala cannot be considered an
international river and that Chile’s past use of the waters should be economically compensated.
See Note N° VRE-DGRB-UAM-009901/2012 from the Ministry of Foreign Affairs of Bolivia to
the General Consulate of Chile in La Paz, 24 May 2012, BCM, Vol. 2, Annex 12. In the same
vein, Note N° VRE-DGRB-UAM-020663/2012 from the Ministry of Foreign Affairs of Bolivia to
the General Consulate of Chile in La Paz, 25 October 2012, CM, Vol. 2, Annex 36, also cited by
Bolivia in this context (BCM, para. 144). All further Notes referred by Bolivia in BCM, p. 89,
footnote 208, dated between 17 January 2013 and 10 April 2014, equally fail to provide any of the
information requested by Chile and affirm that Bolivia’s decision to use the waters of Silala is an
expression of its exercise of full sovereignty. See CM, Vol. 2, Annexes 37.2, 37.4, 37.6, 37.8,
37.10, 37.12 and 38.2.
82 Paragraphs 1 and 2 of Article 17 of the UNWC set forth procedures to be followed in the case of
a reply to a notification regarding planned measures indicating the notified State’s belief that
implementation of planned measures would be inconsistent with Articles 5 or 7, which reflect the
obligations of equitable utilization and prevention of significant harm, respectively. CM, Vol. 2,
Annex 5.
83 UNWC, Article 18, paras. 2 and 3. CM, Vol. 2, Annex 5.
35
27
future work on the Silala, relative to artificial channels and drainage mechanisms
in the exercise of its sovereignty, Bolivia is bound by the rules of international
law set forth above to cooperate with Chile and to provide timely notification of
planned measures that may have an adverse effect on shared water resources, to
exchange data and information and to conduct where appropriate an
environmental impact assessment with respect to such planned work.
2.39 Thus, Bolivia’s first Counter-Claim, that it has the right to decide
whether and how to maintain the channels in its territory, while not disputed by
Chile and therefore outside the jurisdiction of the Court, must be understood in
accordance with Bolivia’s obligation of equitable and reasonable utilization of
Silala waters and the prevention of significant harm, as well as the procedural
obligations set forth above, covered by submissions d) and e) of Chile’s
Memorial.
c. Bolivia’s demand for compensation is unjustified
2.40 It follows from what has been said that Bolivia’s demand for
compensation for any “delivery” to Chile of what Bolivia incorrectly terms
“artificially-flowing waters of the Silala”84 is untenable.
2.41 Bolivia’s demand for compensation is all the more surprising in
view of the fact that it was Bolivia itself, through its authorization of private
company, the FCAB, that unilaterally made the improvements which supposedly
produced an “artificially-enhanced flow.” This demand flies in the face of the
principle that a State is not required to provide compensation for a service that
was not requested or agreed to or, a fortiori, for the results of that service.85 And
84 BCM, paras. 180 and 181 c).
85 This principle has in common the basic reasoning behind those of res inter alios acta alteri
nocere non debet and pacta tertiis nec nocent nec prosunt: one is not bound to something to which
one is not a party. The principle as it applies to treaties is expressed in Article 34 of the Vienna
Convention on the Law of Treaties, 23 May 1969, U.N. Doc. A/CONF.39/27. See generally the
36
28
conversely, an individual or a State providing such an unsolicited service is not
entitled to be compensated for it.
4. The case law, State practice and doctrine referred by Bolivia do not
support the existence of a distinct legal regime for “artificially-enhanced
flow” and are irrelevant to Bolivia’s case
2.42 In Chapter 6 of its Counter-Memorial, “Counter-Claims,” Bolivia
states: “Bolivia’s Counter-Claims are based on the factual and legal conclusions
drawn in the previous Chapters of this Counter-Memorial.”86 The present
subsection will show that the legal bases presented by Bolivia in support of its
contentions concerning “artificially-enhanced flows” are inapposite and in fact
provide no support at all to Bolivia’s claims.
2.43 Bolivia’s second and third Counter-Claims are based on its
argument that States’ obligations to each other in relation to international
watercourses are limited to the “natural flow” of the waters. In support of this
contention Bolivia cites the works of publicists as well as both case and treaty
law. A brief examination of these authorities is all that is necessary to
demonstrate that they are of no help to Bolivia.
2.44 Bolivia first quotes from highly respected authorities in the field of
Public International Law, Oppenheim (Jennings and Watts)87 and Max Huber.88
Separate Opinion of Judge Owada in Question of the Delimitation of the Continental Shelf
Between Nicaragua and Colombia Beyond 200 Nautical Miles from the Nicaragua Coast
(Nicaragua v. Colombia) Preliminary Objections, Judgment, 17 March 2016, I.C.J. Reports 2016,
p. 100, at pp. 174-176.
86 BCM, para. 173.
87 R. Jennings and A. Watts (eds.), Oppenheim’s International Law (Longman, 9th ed., 1996),
p. 585.
37
29
The quotations use the terms “natural” and “naturally” in reference to conditions
of States’ territories, and flow, respectively, but not in contrast to “artificial”
conditions.89 They add no support to Bolivia’s case, and Chile does not disagree
with them.90
2.45 Bolivia next seeks support for its artificial flow theory, upon which
its second and third Counter-Claims are built, in three decisions, the Lake Lanoux
arbitration,91 the Donauversinkung case92 and the Gabčíkovo-Nagymaros Project
case.93 These cases are cited in support of the proposition that “[i]nternational
88 M. Huber, Ein Beitrag zur Lehre von der Gebietshoheit an Grenzflüssen, Zeitschrift für
Völkerrecht und Bundesstaatsrecht, 1907, pp. 29 ff. and 159 ff., translated in S. McCaffrey, The
Law of International Watercourses, Oxford University Press, 2007, p. 132.
89 BCM, para. 80.
90 Bolivia also refers to an article on the present dispute, stating that “[a] manufactured river, in the
form of canals or other man-made systems, would not fall within the rubric of international water
law, since, by definition, such water bodies are proprietary and subject to the agreements that
created them.” BCM, para. 80. The reference given is: “B. Mulligan and G. Eckstein, ‘The
Silala/Siloli Watershed: Dispute Over the Most Vulnerable Basin in South America,’ International
Journal of Water Resources Development, Vol. 27(3), 2011, pp. 595-606.” While confessing some
puzzlement as to exactly what is intended by the authors (the quote seems to refer to a water
conveyance system that is entirely constructed by humans), Chile does not believe the language in
question has anything to do with the Silala watercourse system. The quotation therefore provides
no support for Bolivia’s case.
91 Affaire du Lac Lanoux (Spain v. France), Award of 16 November 1957, Reports of International
Arbitration Awards, Vol. XII, p. 281. Lake Lanoux involved inter-basin transfers of water, using
tunnels and canals, from the Carol River basin to the Ariège River, an equivalent quantity of which
was then transferred back from the Ariège to the Carol, from which the water flowed into Spain.
This was thus hardly a case involving a river’s “natural” flow, at least as Bolivia seems to define
the term.
92 Württemberg and Prussia v. Baden (The Donauversinkung Case), German Staatsgerichtshof,
18 June 1927. The facts of the case have little in common with those of the Silala, involving as
they do the passage of Danube water through the banks and bed of the river during certain periods
of the year, emerging as the source of the Aach River in the Lake Constance/Rhine basin. The
Court found that the resulting “sinking” of the Danube was a natural phenomenon. No “artificial”
flow was involved. Indeed, no works of any kind were involved in the case.
93 Gabčíkovo-Nagymaros Project. It is true that the words “natural flow” appear in the passage
quoted by Bolivia, but Bolivia does not explain how this advances its case. Hungary’s assertion of
a right to 50 per cent of “the natural flow of the Danube” was based on the 1976 Convention cited
by Bolivia, not the judgment in the case. Convention on Regulation of Water Management Issues
of Boundary Waters (Czechoslovakia and Hungary), 31 May 1976, Articles 1-2. Furthermore,
38
30
and domestic judicial decisions […] recognize the legal relevance of the
distinction between the existence of natural and artificial flows.”94 However,
none of these cases involves “artificial” flows, nor uses the term “artificial.” In
addition, none of these cases says anything about, or that could be construed to
support, “the legal relevance of the distinction between the existence of natural
and artificial flows.”95
2.46 Bolivia then attempts to demonstrate that State practice in the form
of treaties shows that at least some agreements “limit their application to the
natural flow of a shared watercourse.”96 For all of the watercourses named (the
Mahakali, Mekong, and Columbia rivers), the scope of the treaties includes
waters affected by substantial dams and other artificial works. Yet none of the
treaties contains the expression “artificial” flows or draws a distinction between
“natural” and “artificial” flows, much less permits an upstream State to demand
compensation for the benefits of its works from downstream States or authorizes
it to assert sovereignty over “artificial” flows. The expression “artificial flows”
instead seems to have been created by Bolivia for the purposes of this case.
2.47 It is therefore not surprising that the State practice on which
Bolivia relies does not differentiate between “artificial” and “natural” flows.97
Bolivia incorrectly assumes that the presence of the expression “natural flows” in
Bolivia omits mention of the fact that the scope of application of the treaty in question includes
artificial works.
94 BCM, para. 81.
95 BCM, para. 81.
96 BCM, para. 82.
97 BCM, para. 82, citing the Treaty Concerning the Integrated Development of the Mahakali River,
India-Nepal, signed on 12 February 1996, 36 I.L.M. 531; Agreement on the Cooperation for the
Sustainable Development of the Mekong River Basin, signed on 5 April 1995, 2069 U.N.T.S. 3;
Treaty Between Canada and the United States of America Relating to Cooperative Development of
the Water Resources of the Columbia River Basin, signed on 17 January 1961, 542 U.N.T.S. 246;
and Treaty Relating to Boundary Waters, and Questions Arising Along the Border between the
United States and Canada, signed on 11 January 1909, 36 Stat. 2448, T.S. No. 548.
39
31
State practice necessarily implies the existence of, what it calls “artificial
flows.”98
2.48 The examples cited by Bolivia that refer to “natural flows” do not
provide for a preferential right to “artificial flows” that is separate from the
regime of international watercourse law. The examples recognize, in some
variation, a right to be protected from significant harm by a reduction, alteration
or obstruction of the “natural flow,” within the context of human development of
an international watercourse. This is true of the Treaty Concerning the Integrated
Development of the Mahakali River Including Sarada Barrage, Tanakpur Barrage
and Pancheshwar Project,99 the Agreement on the Cooperation for the
Sustainable Development of the Mekong River Basin,100 the Columbia River
Treaty,101 and the Boundary Waters Treaty.102
98 BCM, paras. 81-82.
99 BCM, para. 82. Treaty Concerning the Integrated Development of the Mahakali River, India-
Nepal, signed on 12 February 1996, 36 I.L.M. 531. This treaty provides protection against works
that would “adversely affect” (Art. 7) the “natural flow and level” but does not provide for any
preferential ownership of any “artificial flow”. The scope of this treaty also includes the regulation
of the watercourse based on the many artificial works situated on it, as evidenced in the name of
the treaty itself.
100 Agreement on the Cooperation for the Sustainable Development of the Mekong River Basin,
signed on 5 April 1995, 2069 U.N.T.S. 3, Article 6, cited by Bolivia at BCM, para. 82.
101 Treaty Between Canada and the United States of America Relating to Cooperative
Development of the Water Resources of the Columbia River Basin, signed on 17 January 1961,
542 U.N.T.S. 246, referred to by Bolivia at BCM, para. 82. Article XIII provides: “Except as
provided in this Article neither Canada nor the United States of America shall, without the consent
of the other evidenced by an exchange of notes, divert for any use, other than consumptive use,
any water from its natural channel in a way that alters the flow of any water as it crosses the
Canada-United States of America boundary within the Columbia River Basin.”
102 Treaty Relating to Boundary Waters, and Questions Arising Along the Border between the
United States and Canada, signed on 11 January 1909, 36 Stat. 2448, T.S. No. 548. Article II of
the Boundary Waters Treaty does not, as Bolivia states (BCM, para. 82) limit its applicability to
“natural channels” but rather to “waters on [each Party’s] own side of the line which in their
natural channels would flow across the boundary or into boundary waters” and provides remedies
for those affected by changes in the natural channel. The treaty also contemplates the construction
of further works by the parties (Art. III). Thus, regulation of boundary waters is clearly envisaged
in the treaty, which says nothing about “artificial” flows.
40
32
2.49 Bolivia then refers to Article 26 of the UNWC, titled
“Installations,” as support for the propositions that there is “no obligation to
install or to maintain infrastructures for the purposes of increasing the flow and
enhancing the use of transboundary waters. There is no right for a State to require
another State to install or maintain such infrastructures for its benefit.”103 These
are unremarkable propositions, and Chile does not disagree with them. They are
also wholly irrelevant to the circumstances of the case, where Chile never
requested Bolivia to install channels or “enhance” the flow of the Silala River.
2.50 Unable to find any authority in the field of international
watercourses to support its concept of “artificial flows,” Bolivia turns to the law
of the sea. There it finds the term “artificial,” but used in contexts that have
nothing to do with international watercourses, either directly or by analogy.
Instead, the authority in this field cited by Bolivia concerns artificial islands and
maritime delimitation.104 It offers no support for Bolivia’s “artificial flows”
concept.105
2.51 In sum, Bolivia’s second and third Counter-Claims assume that
under international law there is a distinct legal regime relating to an “artificially
enhanced flow” of an international watercourse; Bolivia however, has failed to
103 BCM, para. 83.
104 BCM, paras. 85-90.
105 Unsurprisingly, the United Nations Convention on the Law of the Sea (UNCLOS) treats
artificial structures and natural features differently. For example, artificial islands, installations and
structures in the exclusive economic zone “have no territorial sea of their own […].” United
Nations Convention on the Law of the Sea, 3 December 1982, 1833 U.N.T.S. 3, Article 60,
para. 8. Bolivia’s reliance on the South China Sea Arbitration (The Republic of Philippines v. The
People’s Republic of China), PCA Case No. 2013-19, Award of 12 July 2016; Maritime
Delimitation in the Black Sea (Romania v. Ukraine), Judgment, I.C.J. Reports 2009, p. 61;
Maritime Delimitation and Territorial Questions between Qatar and Bahrain, Merits, Judgment,
I.C.J. Reports 2001, p. 40; Fisheries case (United Kingdom v. Norway), Judgment, I.C.J. Reports
1951, p. 116; and Land and Maritime Boundary between Cameroon and Nigeria (Cameroon v.
Nigeria: Equatorial Guinea intervening), Judgment, I.C.J. Reports 2002, p. 303, is equally
unavailing. Indeed, Bolivia fails to explain how they are relevant, or apposite, to the present case.
41
33
cite any relevant authority supporting the existence of such a regime. The
doctrine, treaty practice and case law Bolivia refers to fail entirely to address the
topic. The fact that Bolivia’s theory is without precedent and unsupported by any
source of international law leads to the unavoidable conclusion that Bolivia’s
theory, upon which its counter-claims are constructed, is just that, and has no
legal value.
B. The historical background relevant to Bolivia’s counter-claims: key
omissions by Bolivia
2.52 In formulating its counter-claims, Bolivia has elected to pass over a
series of key facts. First, it ignores its own century-long practice recognising the
Silala River as a transboundary watercourse, a practice that was not accompanied
by any statements as to there being a distinction between “natural” and
“artificial” flow. This omission is considered further in section 1 below. Second,
as detailed in section 2, Bolivia passes over the fact that the waters of the Silala
River in Chilean territory were licensed by Chile in 1906 to the British private
company FCAB prior even to the Bolivian concession of 1908, and likewise the
fact that the construction of the channels in Bolivia took place in the late-1920s to
improve the quality, not quantity, of the water. These are important (but omitted)
facts because they show how the flow of the waters of the Silala into Chile was
not, and is not, dependent on the excavated earth channels and lining developed
by FCAB on which Bolivia has now chosen to build a case. For many years, the
waters were considered capable of exploitation, and were exploited, without the
FCAB’s channels.
2.53 This section is completed by examining two further key omissions
by Bolivia: its failure to identify that the channels now at issue were constructed
on Bolivian territory with Bolivian authorization, and are therefore a consequence
42
34
of Bolivia’s own sovereign acts (section 3); and the absence of any explanation
by Bolivia of why it has not simply removed the channels and restored the
wetlands in Bolivia (section 4).
1. Bolivia ignores almost 100 years of joint Bolivian-Chilean recognition of
the Silala as a river without distinguishing “natural”
from “artificial” flow
2.54 In its Memorial, Chile demonstrated the almost century long
recognition by both States of the Silala as a natural river and transboundary
watercourse.106 See, for example:
(a) The depiction of the Silala River on Chilean and Bolivian
cartography, including the Map appended to the 1904 Treaty of
Peace and Amity signed by the representatives of both
States.107 As recently as 1997, the Silala River was depicted on
the Geological Map of Bolivia by the Bolivian Geological
Survey (SERGEOMIN)108 and on the official Map of the area
prepared by the Bolivian Military Geographical Institute
(I.G.M.).109 During all this time, Bolivia never referred to, or
distinguished between, “artificial” and “natural” flows.
(b) The consistent recognition of the Silala River by the Chilean
and Bolivian members of the various Mixed Commissions in
charge of the demarcation and revision of the international
106 CM, paras. 4.13-4.35.
107 Map Appended to the Treaty of Peace and Amity, 20 October 1904. CM, Vol. 6, Annex 82.
108 Bolivian Geology and Mining Survey (SERGEOMIN), Geological Map of Bolivia, Sheet 5927-
6027 Silala-Sanabria, ed. March 1997. CM, Vol. 6, Annex 89.
109 Bolivian Military Geographical Institute (I.G.M.), Map of South America (Bolivia) Volcán
Juriques, 1st ed., reissued May 1997. CM, Vol. 6, Annex 90.
43
35
boundary, from 1906 until the late 1990s.110 Even after
Bolivia’s abrupt change of position in 1999, Bolivian domestic
legislation and its submissions before the Secretariat of the
Ramsar Convention continued to refer to the Silala as a river.111
Again, no reference was ever made to “natural” versus
“artificial” flows.
2.55 Bolivia has made no attempt to engage with this evidence. It asserts
that “Chile’s Memorial relies on inaccurate interpretations of Bolivian
cartography, minutes, and statements regarding the Silala waters”.112 Yet no
explanation is given for the numerous representations and descriptions of the
Silala River by the Bolivian authorities. Bolivia also asserts that, at the time the
multiple documents were produced, “both States lacked sufficient scientific
evidence to accurately determine the nature of the Silala waters.”113 This is an
irrelevance, and also incorrect:
(a) No scientific evidence is necessary to confirm that the Silala is
a transboundary watercourse, whose waters flow naturally from
Bolivia into Chile. In the words of Bolivia’s Chair of the
Bolivian Boundary Commission and President of the Mixed
Boundary Commission in 1996, who visited the area many
times together with his Chilean counterpart:
“It rises from two main springs and receives additional
waters from other minor springs. The narrow riverbed
that is formed, called Silala, runs approximately two
kilometers through Bolivian territory before it crosses the
110 CM, paras. 4.36-4.55.
111 CM, paras. 4.62-4.66.
112 BCM, para. 25.
113 BCM, para. 25.
44
36
boundary at a point of the east-west slope of the glen
between Cerro Inacaliri and Cerro Silala. The inclination
of the terrain has been established by experts to be
around 30% [sic, more likely 3%], its river bed is narrow
and its crystalline waters follow the course that, due to
the force of gravity, goes downhill into Chilean
territory.”114
(b) Following Bolivia’s change of position in 1999, Chile did agree
(in 2004) to form a Technical Commission and conduct joint
technical and scientific studies on the Silala River, but this was
not because it lacked scientific evidence, as Bolivia suggests.115
Rather, Chile acted in the reasonable expectation that Bolivia,
once faced with the incontrovertible facts, would revert to its
pre-1999 position of acknowledging the Silala as an
international river, as it has now done.
2.56 Bolivia also contends that the expert reports submitted by both
Parties confirm Bolivia’s position that the Silala constitutes an “artificiallyenhanced
watercourse”.116 This is incorrect. Neither Chile’s nor Bolivia’s experts
make any mention of an “artificially-enhanced watercourse” or distinguish
“artificial” from “natural” flows. These are concepts of Bolivia’s own invention.
2.57 In short, in defence to the counter-claims, Chile reiterates the
evidence presented in its Memorial, not addressed by Bolivia, demonstrating
Bolivia’s century-long recognition of the Silala River as an international
watercourse, without making any distinction between “natural” and “artificial”
flows.
114 Presencia, “Dialogue on Friday with Dr. Teodosio Imaña Castro”, La Paz, 31 May 1996. CM,
Vol. 3, Annex 71.
115 BCM, para. 25.
116 BCM, para. 42.
45
37
2. The true facts with respect to the 1906 and 1908 concessions and the later
(1928) channelization in Bolivia for sanitary reasons
2.58 Bolivia makes no mention of the 1906 Chilean concession of the
Silala waters, obtained by the British private company FCAB from the Chilean
authorities.117 This 1906 Chilean concession matters because it shows that in
1906, prior to the construction of any waterworks in either Chile or Bolivia, the
Silala River was flowing naturally across the border from Bolivia into Chile. Had
it been otherwise, Chile would not have been in a position to grant a concession
with respect to the waters.
2.59 Bolivia contends that, prior to the channelization, the waters of the
Silala did not flow naturally across the border “in the rate and volume adequate
for the Railway Company’s intended purpose”.118 Bolivia relies on the language
of the 1908 Bolivian deed of concession, stating that “[b]y building intake and
channeling works, the previously mentioned springs could be used, even if at
increased cost […]”,119 and that “the projected work shall make usable waters
that are currently being lost benefitting no one.”120
2.60 However, Bolivia’s account of the construction of the waterworks
and channels is chronologically inaccurate, ignores the evidence on record and is
highly misleading.
117 Deed of Concession by the State of Chile of the Waters of the Siloli (N° 1.892) to The
Antofagasta (Chili) and Bolivia Railway Company Limited, 31 July 1906 (henceforth “1906
Chilean concession”). CM, Vol. 3, Annex 55. See for its discussion, CM, paras. 2.21, 4.56-4.58;
cf. BCM, para. 48.
118 BCM, para. 65; see also, para. 63.
119 BCM, paras. 63 and 65, quoting from Deed of Concession by the State of Bolivia of the Waters
of the Siloli (N° 48) to The Antofagasta (Chili) and Bolivia Railway Company Limited,
28 October 1908. CM, Vol. 3, Annex 41.
120 BCM, para. 67, quoting from Deed of Concession by the State of Bolivia of the Waters of the
Siloli (N° 48) to The Antofagasta (Chili) and Bolivia Railway Company Limited, 28 October
1908. CM, Vol. 3, Annex 41.
46
38
2.61 As Chile explained in its Memorial, the first works were carried out
in 1910, with an intake (Intake N° 1) being built in Bolivian territory just
600 metres from the international boundary.121 There is no indication that any
channels were built in the Bolivian wetlands or in the Silala ravine at that time
and Bolivia has not shown otherwise.
2.62 The correct position is that the channelization was undertaken in
1928, almost eighteen years after FCAB started to take the Silala waters, for
sanitary reasons.122 The reason for constructing the channels is well documented
in correspondence between the General Manager of FCAB in Antofagasta and the
Board of Directors of FCAB in London, as follows:
“For some time past a little difficulty has been encountered in keeping
the water from this source up to that high standard of purity desired,
and suspicions have been aroused by the fact that certain eggs of fly
have been discovered, under microscopic examination, in the water in
Antofagasta. These eggs hatch out into a specie of small green fly. The
cause was finally traced to the head works in the Siloli valley where
there is considerable vegetable growth through which the water has to
flow before reaching the intake.”123
2.63 The General Manager discussed two possible schemes to solve this
problem:
“The schemes for overcoming this difficulty have been prepared by the
Waterworks Engineer, the first, that of cleaning up the course of the
water through the valley by cutting an earth channel from the upper
121 CM, para. 4.60. See for FCAB’s formal request to the Bolivian authorities to introduce the
pipelines into Bolivia through Chilean territory, Request from FCAB to the Government of
Bolivia, 3 August 1910, CM, Vol. 3, Annex 65; and for Bolivia’s authorization thereof,
Communication N° 71 from the Government of Bolivia to The Antofagasta (Chili) and Bolivia
Railway Company Limited, 9 August 1910, CM, Vol. 3, Annex 42.
122 CM, para. 4.61.
123 Letter from the General Manager of FCAB in Chile to the Secretary of the Board of Directors
of FCAB in London, 27 January 1928. CM, Vol. 3, Annex 67.1.
47
39
springs to the existing intake works, and also a branch trench from the
“Cajon” springs near the intake. The second scheme provides for the
construction of a concrete channel in place of the earth channel.”124
2.64 Given the well-documented historical record of the 1928
channelization, it is untenable for Bolivia to say that “Chile ignores the very
purpose and justification for the construction of the channels”.125 To the contrary,
the historical record does not support Bolivia’s position that the channels were
built to increase the water supply.126
2.65 Bolivia’s current assertions that “prior to the installation of the
artificial channels, Silala waters within Bolivia’s bofedales region was relatively
stagnant, with a considerably reduced cross-border water flow on the surface as
compared to the present”,127 and that in its pre-channelized condition the Silala in
Bolivia “did not flow naturally across the border in the manner, rate and volume
that met the needs of the Railway Company”,128 are incorrect.
124 Ibid.
125 BCM, para. 58.
126 A fortiori, there is no evidence that the channelization in Orientales and Cajones “intentionally
depleted those fragile wetlands” as claimed by Bolivia (BCM, para. 69). In fact, the FCAB opted
for earth instead of concrete channels, due to the urgency of the matter and without giving any
consideration to the increased efficiency that concrete channels may provide, see Letter from the
General Manager of FCAB in Chile to the Secretary of the Board of Directors of FCAB in
London, 27 January 1928: “The whole matter has been explained to Mr. Bolden during his visit
with a recommendation that the work of renewing and improving the existing intake works and the
cutting of the earth channel between the present springs and the intake of the Siloli pipe line
should be carried out forthwith in view of the urgent necessity of same. It was explained that in the
event of it being necessary subsequently to construct a concrete channel, the expenditure incurred
in the cutting of the earth channel would then form a preliminary work for the concrete channel.”
CM, Vol. 3, Annex 67.1. As far as Chile is aware, no concrete channels were constructed
afterwards and the 1928 earth channels are still more or less in place today.
127 BCM, para. 62.
128 BCM, para. 65.
48
40
2.66 The 1906 Chilean concession was requested to provide “abundant
potable water of good quality to the city of Antofagasta”,129 and was considered
significant enough to “solve […] the issue of potable water supply for the
aforementioned city”.130 This demonstrates that there was a considerable flow
entering from Bolivia into Chile prior to any waterworks or channelization.
2.67 It is also incorrect, as Bolivia now says, that a second pipeline
(Pipeline N° 2) was constructed in 1942 to convey the waters “generated by the
channelization” into Chilean territory.131 As early as in 1916, a second pipeline
had been planned, but FCAB had not been able to raise the necessary capital.132
The capacity of the only pipeline installed at that time (Pipeline N° 1) was
approximately 75 l/s,133 and a second pipeline would have allowed the FCAB to
capture double that amount.134 Pipeline N° 2 was eventually constructed in 1942,
fourteen years after the excavation of the channels in 1928, and bore no relation
whatsoever to that earlier set of works.
2.68 Chile notes in passing that Bolivia asserts the use of explosives at
the Silala headwaters to remove soil and rocks, again to increase the discharge of
the springs.135 Bolivia does not provide any evidence for this, other than a picture
129 1906 Chilean Concession. CM, Vol. 3, p. 201.
130 Ibid., p. 205.
131 BCM, para. 52.
132 Letter from the General Manager of FCAB in Chile to the Secretary of the Board of Directors
of FCAB, 7 April 1916. Chile’s Reply (“CR”), Vol. 2, Annex 92. See also Letter from the
General Manager of FCAB in Chile to the Secretary of the Board of Directors of FCAB,
8 September 1916, explaining that the second pipeline was considered necessary to satisfy
increasing water demands from American mining company “Chile Exploration Company”
(Chilex). CR, Vol. 2, Annex 93.
133 Robert H. Fox, The Waterworks Department of the Antofagasta (Chili) & Bolivia Railway
Company, South African Journal of Science, 1922, p. 124. CM, Vol. 3, Annex 75. See also CM,
para. 2.22.
134 Letter from the General Manager of FCAB in Chile to the Secretary of the Board of Directors
of FCAB, 7 April 1916. CR, Vol. 2, Annex 92.
135 BCM, paras. 59 and 61.
49
41
of allegedly blasted rocks arranged along an unidentified section of the Silala
River.136 From the image presented, it is impossible to know whether, where,
when, why and by whom these rocks were blasted. It proves nothing.137
3. Bolivia’s failure to take account of the simple fact that the channels were
built with Bolivian authorization
2.69 Even if Bolivia were right that the channels were constructed to
increase spring discharge in the Silala wetlands (it is not), this would be an
irrelevance. The channels were constructed by British private company FCAB
(not Chile) pursuant to the 1908 Bolivian concession, and therefore with Bolivian
authorization.138
2.70 The 1908 Bolivian concession is a sovereign act of the Bolivian
State, regulated by Bolivian domestic law. Its conditions and enforcement cannot
possibly lead to international responsibility on the part of Chile, as indeed both
Parties have confirmed on various occasions in the past. For example, by Note of
3 September 1999, directed to Chile, Bolivia stated:
“It is worth emphasizing that said concession was granted by the
Prefecture of the Department of Potosí to a private Company and not to
136 BCM, p. 48, Fig. 19.
137 Chile calls attention to Bolivia’s reference to a case study on blasting as a method to increase
yield from wells by a factor of 6 to 20, suggesting that such techniques may have been used at the
Silala to enhance water flow from the springs. See BCM, para. 61, with reference to F.G. Driscoll,
“Blasting – It turns Dry Holes into Wet Ones”, Johnson Drillers’ Journal, Nov/Dec, 1978,
Johnson Division, UOP, Inc. St. Paul, MN, p. 3. The referenced article describes fracturing
techniques at depth, bearing no relation whatsoever to the situation at the Silala, where springs
discharge naturally from a shallow aquifer. Chile’s experts consider that significant development
of spring flow is not possible using these blasting methods. Peach, D.W. and Wheater, H.S.,
Concerning the Geology, Hydrogeology and Hydrochemistry of the Silala River Basin (henceforth
“Peach and Wheater (2019)”), pp. 52 and 54.
138 Deed of Concession by the State of Bolivia of the Waters of the Siloli (No. 48) to The
Antofagasta (Chili) and Bolivia Railway Company Limited, 28 October 1908, p. 19. CM, Vol. 3,
Annex 41.
50
42
the Chilean State. Hence, all actions undertaken to date, as well as those
that the cited Company carried out, were in the private sphere and with
full acknowledgement of the Bolivian jurisdiction.”139
2.71 Bolivia had the right to authorize the construction of the channels
by the FCAB in Bolivian territory. But, if those channels had any impact on the
cross-boundary surface flow of the Silala River (they have no significant impact),
that could not somehow be attributed to, or lead to an obligation to pay
compensation on the part of, Chile.
4. Notwithstanding the termination in 1997 of the 1908 concession, Bolivia
has not removed the channels and restored the wetlands
2.72 Bolivia contends that the wetlands in Bolivia have been adversely
affected by the construction of the waterworks in its territory.140 To restore the
wetlands, Bolivia says that it may have to modify the channels and drainage
mechanisms.141 To this end, it asks the Court to declare that it has sovereignty
over these installations and the right to decide whether and how to maintain them
(Counter-Claim a)).142
2.73 There is no need (and no basis) for any such declaration. Chile has
no objection to Bolivia’s sovereign decision to restore the wetlands in its
territory, without prejudice to Bolivia’s obligations towards Chile under
customary international law and Chile’s right to equitable and reasonable
139 Note N° GMI-656/99 from the Ministry of Foreign Affairs of Bolivia to the General Consulate
of Chile in La Paz, 3 September 1999. CM, Vol. 2, Annex 27.
140 BCM, paras. 73 and 176, quoting from Ramsar Convention Secretariat, Report Ramsar
Advisory Mission N° 84, Ramsar, Site Los Lipez, Bolivia, 2018 (henceforth “Ramsar Report”).
BCM, Vol. 5, Annex 18.
141 BCM, para. 179.
142 BCM, para. 181 a).
51
43
utilization of the waters of the Silala. There is no dispute between the Parties in
this regard.
2.74 Moreover, Bolivia’s interest in the restoration of the wetlands in its
territory appears to be a matter of secondary importance to it. As follows from its
second and third Counter-Claims, Bolivia would be willing to “deliver” to Chile
the “artificially-flowing” waters of the Silala that are “engineered, enhanced, or
produced in its territory”, against payment of compensation by Chile to be agreed
upon. Such “delivery” would, on Bolivia’s case, apparently be premised on
maintaining the channelization in the Bolivian wetlands (according to Bolivia,
this produces the “artificial” flow for which payment is said to be due), rather
than restoring the wetlands to their natural condition.143
2.75 Indeed, Chile notes that Bolivia could have removed the channels
and restored the wetlands at any time over the last century, or at least following
the termination of the 1908 Bolivia concession in May 1997.144 This was not,
however, the course of action taken by Bolivia:
(a) Instead, in April 2000, Bolivia granted a new concession to the
waters of the Silala, this time to Bolivian company DUCTEC
S.R.L., for the duration of forty years. This new concession
authorized the commercialization and exportation of the waters for
industrial use and human consumption.145
143 BCM, para. 181 b) and c).
144 Administrative Resolution N° 71/97 by the Prefecture of the Department of Potosí, 14 May
1997. CM, Vol. 3, Annex 46.
145 Concession Contract for the Use and Exploitation of the Springs of the Silala Between the
Bolivian Superintendent of Basic Sanitation and DUCTEC S.R.L., 25 April 2000. CM, Vol. 3,
Annex 48. The concession excluded the use of the waters of the Silala for potable water and
sewerage services in Bolivia without an additional public utility concession, as well as for mining
activities by third parties in Bolivian territory. Hence, the only potential end-users of the water
rights granted to DUCTEC were in Chile. DUCTEC attempted to invoice FCAB and Corporación
Nacional del Cobre de Chile (CODELCO) for their use of the water on Chilean territory, to no
52
44
(b) In the years following, Bolivia considered and partly developed
several projects to make use of the waters of the Silala River,
including a fish farm, a small dam and a mineral water bottling
plant.146 When requested by Chile to provide information on these
projects in accordance with the international law applicable to
international watercourses,147 Bolivia affirmed its full sovereignty
over the use and exploitation of these resources.148
2.76 Bolivia’s newly stated intention to restore the wetlands appears to
have coincided with Chile’s submission of its Application on 6 June 2016.
Shortly after, on 6 July 2016, Bolivia met with the Ramsar Secretariat in Geneva
in relation to the Ramsar site Los Lípez of which the Silala forms part.149 This
meeting was followed by Bolivia’s request, by letter of 27 July 2016, for a
Ramsar Mission on site, expressing its “concern about the negative changes
observed as to the ecological characteristics of the Los Lipes (sic) site, the Silala
avail. See Invoice N° 003/00 from DUCTEC to CODELCO, 5 May 2000. CM, Vol. 3, Annex 76.
See for Chile’s formal objection to Bolivia’s concession of the totality of the waters of the Silala to
DUCTEC, Note N° 006738 from the Ministry of Foreign Affairs of Chile to the Ministry of
Foreign Affairs of Bolivia, 27 April 2000. CM, Vol. 2, Annex 31. The DUCTEC concession was
terminated on 30 May 2003, due to the illegitimacy of the Concession contract, see Bolivian
Administrative Resolution Nº 75/2003 by the Superintendency of Basic Sanitation, 30 May 2003.
CM, Vol. 3, Annex 50.
146 CM, para. 3.26.
147 Note N° 199/39 from the General Consulate of Chile in La Paz to the Ministry of Foreign
Affairs of Bolivia, 7 May 2012. CM, Vol. 2, Annex 34. Note N° 389/149 from the General
Consulate of Chile in La Paz to the Ministry of Foreign Affairs of Bolivia, 9 October 2012. CM,
Vol. 2, Annex 35.
148 See, among others, Note N° VRE-DGRB-UAM-020663/2012 from the Ministry of Foreign
Affairs of Bolivia to the General Consulate of Chile in La Paz, 25 October 2012. CM, Vol. 2,
Annex 36.
149 Note N° VRE-Cs-58/2016 from the Ministry of Foreign Affairs of Bolivia to the Senior
Advisor for the Americas of the Ramsar Convention Secretariat, 27 July 2016. CR, Vol. 2,
Annex 97.
53
45
wetland and related areas, caused by the artificial channelization of its springs for
the purpose of exploiting those waters […]”.150
2.77 In any event, Bolivia has not explained to the Court why it has not
restored the wetlands, and Chile wishes to emphasise that it has not been
responsible (and could not have been responsible) for delay with respect to a
restoration on Bolivian territory. Chile also considers that the full restoration of
the wetlands would have minimal impact on the cross-border flow into Chile, and
encourages Bolivia to undertake such measures as are necessary and appropriate
to the wetland restoration (whilst complying with its obligations to Chile,
including by way of notification and consultation).
C. Conclusion: The distinction between “natural” and “artificiallyenhanced”
flow with the legal consequences alleged by Bolivia is untenable
under international law and Bolivia’s second and third Counter-Claims
must be dismissed
2.78 Bolivia’s Counter-Claims b) and c) are premised on a distinction
made by Bolivia between “natural flow” and “artificially-enhanced flow” that has
no support in international law and indeed goes counter to the accepted principle
of equitable and reasonable utilization of international watercourses. Bolivia also
ignores key elements of the history of the Silala River and its uses and passes
over the basic fact that the channelization of the Silala was built in Bolivia, with
Bolivian authorization, and could have been removed by Bolivia long ago to
restore the wetlands in its territory. Bolivia’s second and third Counter-Claims
have no justification and must be dismissed.
150 Ibid. Chile notes that in 2015, Bolivia reported “no negative change” in any of its Ramsar sites
to the Ramsar Secretariat, see National Report on the Implementation of the Ramsar Convention
on Wetlands submitted by the Plurinational State of Bolivia to the 12th Meeting of the Conference
of the Contracting Parties, 2 January 2015, response to question 2.6.2. CR, Vol. 2, Annex 95.

55
47
CHAPTER 3
BOLIVIA’S CONTENTIONS ON THE ALLEGED IMPACT OF THE
CHANNELIZATION IN BOLIVIA ARE UNTENABLE AS A MATTER OF
FACT
3.1 Bolivia’s second and third Counter-Claims are based on a series of
contentions as to the impacts of the 1928 channelization in Bolivia. Those
contentions are based on the estimates of Bolivia’s experts, DHI, arrived at
through a modelling exercise that DHI has carried out for the purposes of the
current case. In this Chapter, Chile considers these factual contentions, but recalls
its position that Bolivia’s second and third Counter-Claims can and should be
dismissed solely on legal grounds.
3.2 As is discussed in section A below, the Parties are largely in
agreement on the nature and functioning of the Silala River as an international
watercourse.
3.3 As explained in section B, the alleged impact of the channelization
in Bolivia on the surface flow of the Silala (estimated by Bolivia at 30-40%
“artificially-enhanced flow”) is grossly overstated, if indeed such impact exists at
all.
3.4 Chile has requested from Bolivia certain data and information that
Chile’s expert Prof. Wheater considers necessary to fully understand and
critically assess the DHI model and its results.151 Bolivia has not provided this
information in time to be considered by Chile’s experts in the present
151 Notes from the Agent of the Republic of Chile to the Agent of the Plurinational State of Bolivia
of 5 November 2018 (CR, Vol. 2, Annex 99.1), 30 November 2018 (CR, Vol. 2, Annex 99.3) and
21 December 2018 (CR, Vol. 2, Annex 99.5).
56
48
submission.152 Chile affirms its right to refer to the requested data and
information once these have been reviewed and analyzed by Chile’s experts.
3.5 Nevertheless, the information on the record is sufficient to make
the following key points, that will be developed in this Chapter:
(a) Independently of any modelling efforts, all water in the Silala River
basin will flow from Bolivia into Chile, whether as surface water
or groundwater;
(b) The inflow in each scenario modelled by DHI is different, causing
the outflow in each scenario to be different as well, proving
nothing about the impact of channelization;
(c) DHI’s modelling results are entirely dependent on magnifying the
impact of the channels by modelling just 1% of the relevant area
(the Near Field);
(d) Bolivia relies on a 2018 Report of the Ramsar Convention
Secretariat153 and its contentions that the wetlands in Bolivia are
severely deteriorated, however this is contradicted by Bolivia’s
own 2017 Castel study154 and a 2016 Ministerial report.155
152 Notes from the Agent of the Plurinational State of Bolivia to the Agent of the Republic of Chile
of 22 November 2018 (CR, Vol. 2, Annex 99.2), 11 December 2018 (CR, Vol. 2, Annex 99.4),
11 January 2019 (CR, Vol. 2, Annex 99.6) and 7 February 2019 (CR, Vol. 2, Annex 99.7).
153 Ramsar Report. BCM, Vol. 5, Annex 18.
154 Ana Paola Castel, Multi-Temporal Analysis through Satellite Images of the High Andean
Wetlands (bofedales) of the Silala Springs, Potosí – Bolivia, September 2017 (henceforth
“Castel”). CR, Vol. 2, Annex 98.
155 Ministry of the Environment and Water of Bolivia, Characterization of Water Resources in the
Southwest of the Department of Potosí – Municipality of San Pablo de Lipez “Wetlands of Silala
Valley and Adjacent Sectors” (Volume II), July 2016. CR, Vol. 2, Annex 96.
57
49
3.6 In short, there is nothing close to a sound factual and scientific
basis for Bolivia’s second and third Counter-Claims, which must therefore be
rejected.
A. Chile and Bolivia largely agree on the nature and functioning of the
Silala River as an international watercourse
1. Chile and Bolivia agree that the Silala River is a perennial flow that rises
at two sets of springs in Bolivia and flows along the natural topographic
gradient from Bolivia into Chile
3.7 There is agreement between the Parties that the Silala River is a
complex groundwater-surface water system that originates in two sets of springs
in Bolivia and crosses the international border from Bolivia into Chile, due to the
natural topographical gradient.156 The gradient is estimated by Bolivia at
approximately 3.7%.157 The Parties also agree that the channelization in Bolivian
territory did not alter or divert the natural direction of the flow of the water from
Bolivia towards Chile.158 Thus the direction of the flow of Silala River waters has
been the same for thousands of years.
3.8 The Parties agree that surface water runoff contributes a very minor
proportion of the average daily flow of the Silala River, which is groundwater
dominated.159 They also agree that the Silala River interacts with groundwater
throughout its course and that the direction of the subsurface water (as of the
156 BCM, paras. 41-44.
157 BCM, para. 44.
158 DHI Report. BCM, Vol. 2, p. 267: “14. The canals have changed the amount of discharge from
the Silala springs but not the direction of natural outflow from the Silala wetlands. Also, in a
situation without the canals, the discharge direction is towards Chile.” (Emphasis in the original).
159 BCM, para. 47.
58
50
surface water) is westward towards and into Chile.160 Bolivia’s expert DHI
estimates that the groundwater flow is at least of the same order of magnitude as
the surface flow.161
2. Chile and Bolivia agree that the 1928 channelization in Bolivia has only a
minor effect on the direct loss of water to evaporation of no more than 2%
of the current cross-border flow
3.9 In addition, Chile and Bolivia agree that the channelization
undertaken in 1928 may have resulted in reduced direct loss of water to
evaporation, due to a possible reduction of the extent of surface water in the
Bolivian wetlands. In Chile’s Memorial, Chile’s experts estimated the reduced
evaporation at 1.3 l/s or 0.7% of the cross-border flow, but given the uncertainty
in this calculation, they also included a more conservative estimate of 3.4 l/s or
2% of the cross border flow.162 DHI’s estimate is slightly lower than that, at 2 to
3 l/s of the combined cross-border groundwater and surface flows.163 Both sides
agree that this reduction of evaporation is a small component of the total water
balance of the Silala River system and it of course does not account for the
30-40% “artificially-enhanced” flows claimed by Bolivia.
3.10 The conclusions as to evaporation have been reinforced by recent
studies by Chile, in which estimates of evaporation from Bolivia’s wetlands (with
channelization) are very similar to evaporation from a similar wetland in the
160 BCM, para. 47. See also DHI Report, BCM, Vol. 5, p. 84: “groundwater level gradients and
hydrogeological properties clearly indicate groundwater flow from Bolivia to Chile […]”.
161 DHI Report. BCM, Vol. 5, p. 84.
162 CM, Vol. 1, p. 133.
163 DHI Report. BCM, Vol. 2, p. 267.
59
51
Silala River basin in Chile (which has no channelization).164 This suggests that
the effects of channelization on the water balance, if any, are very limited.
3. Chile and Bolivia agree on the complexity of the groundwater flow
systems of the Silala, having different origins and recharge areas
3.11 Chile’s and Bolivia’s experts agree that the springs in the
Orientales and Cajones wetlands in Bolivia have different isotopic and chemical
compositions, implying different origins and different recharge areas.165 On the
basis of data provided by Bolivia, the springs at the Orientales (Southern)
wetland have chemical similarities to the deeper groundwater analysed in Chile
and are likely a mix of locally recharged groundwater and groundwater from a
regional aquifer. The spring waters at the Cajones (Northern) wetland have a
chemical composition similar to that of springs found on the northern side of the
Silala River ravine in Chile and, like them, have a locally recharged origin.166
Hence, both Chile’s and Bolivia’s experts confirm the complex nature of the
Silala River and groundwater flow systems.
164 In fact, estimated evaporation from the Bolivian wetlands with channelization is 10% greater
than from the Chilean wetland without channelization, but this is within the margin of error for the
method used. Wheater and Peach (2019), p. 41.
165 DHI Report. BCM, Vol. 4, p. 103.
166 Peach and Wheater (2019), p. 46. While it is likely that the springs have different ages, the
dates provided by Bolivia’s expert are incorrect, see Peach and Wheater (2019), pp. 45-46.
60
52
4. While Chile and Bolivia maintain different interpretations of the geology
and hydrogeology of the Silala River basin, this does not affect their
common understanding of the nature of the Silala River as an
international watercourse
3.12 Despite these important convergences between Chile’s and
Bolivia’s experts, they maintain different interpretations of the geology and
hydrogeology of the Silala River basin.
3.13 Bolivia’s proposed succession and dates of (permeable) ignimbrite
and lava deposits in the Silala River valley cannot be reconciled with Chile’s
recent geological mapping, radiometric dating results, drilling evidence and
pumping test results.167 This means that the aquifer system in the ignimbrites
identified in Chile has not been recognized by Bolivia.168 On the other hand,
Bolivia infers a massive geological fault system that would run from the
Orientales wetland to the Cajones wetland in Bolivia, bending around and
following the line of the Silala River into Chile, that Chile’s experts consider
highly implausible. This inferred fault is not evidenced by any displacement of
rocks on either side of the river valley, as would necessarily occur in a major fault
zone.169
3.14 These differences in interpretation do not affect Chile’s and
Bolivia’s common understanding of the Silala River as an international
watercourse. However, they do affect the reliability of the DHI Near Field model,
which is the only source of support for Bolivia’s claims for the large effects of
channelization, as discussed in more detail in section B below.
167 Peach and Wheater (2019), p. 52.
168 Peach and Wheater (2019), p. 7.
169 Peach and Wheater (2019), pp. 22-23, 30-31 and 34.
61
53
B. Bolivia’s estimation of the impact of the 1928 channelization in
Bolivia on the cross-boundary surface flows (30-40% “artificial flow”) is
untenable and based on a fundamentally flawed numerical model
1. The three scenarios (“Baseline”, “No Canal” and “Restored Wetlands”)
used by Bolivia to calculate the 30-40% “artificial flow” are inconsistent
with the law of conservation of mass and cannot lead to a reliable
calculation
3.15 The assessment of Bolivia’s experts, DHI, of the impact of
channelization is based entirely on a model that they have developed. This is an
integrated (surface water and groundwater) numerical model of the Silala, within
an area called the Silala Near Field. The Silala Near Field, however, covers an
area of just 2.56 km2 from the international border to just upstream of the Cajones
and Orientales wetlands. This corresponds to just 1.1% of the total Silala
groundwater catchment of 234.2 km2,170 also referred to as the Silala Far Field, as
can be seen on Figure 1. The use of such a small area, as further identified below,
leads to wholly unreliable results so far as concerns the DHI modelling exercise.
170 DHI Report. BCM, Vol. 2, p. 289.
62
54
Figure 1. Approximate extents of the Silala Near Field (reproduced from DHI Report.
BCM, Vol. 2, p. 271, Figure 3).
3.16 The Near Field model has been run by DHI for different scenarios,
with and without channelization, and with and without a layer of restored peat.
This is said to allow for assessment of the effects of the channel and drainage
network in Bolivia on the surface and groundwater flows.171
(a) The Baseline Scenario of the DHI Near Field model represents the
current situation, with the existing channels in Bolivia in place.
(b) The No Canal Scenario represents a situation in which these
channels are removed and the surface water flow is largely
controlled by the surface topographical slope.
(c) Finally, the Restored Wetlands Scenario considers how the fully
restored wetlands might be expected to function in a distant future
171 DHI Report. BCM, Vol. 5, p. 66.
International border
Silale canal
63
55
by considering the possibility of long-term peat accumulation in
the wetlands.172
3.17 The Scenario results are shown in Table 1 of Annex H to the DHI
Final Report173:
Table 1. DHI’s results of its modelling of different scenarios (reproduced from DHI
Report. BCM, Vol. 5, p. 67, Table 1).
3.18 As Chile’s experts point out, the Near Field model is severely
flawed, in several important respects. In particular, the exaggerated effect of the
channelization is largely driven by incorrectly defined boundary conditions of the
172 DHI Report. BCM, Vol. 5, p. 66.
173 DHI Report. BCM, Vol. 5, p. 67.
Table 1 Summary of key scenario results
Baseline Scenario No canal scenario Restored wetlands
Water balance
Volume Flow Volume Flow Volume Flow
component
equivalent equivalent equivalent equivalent equivalent equivalent
(mm/y) (Vs) (mm/y) (1/s) (mm/y) (Vs)
Inflow 3116 253 2722 221 2655 216
Storage change 49 4 12 1 64 5
Evapotranspiration 125 10 150 12 164 13
Error 25 2 0 0 -2 0
Outflow (canals) 1846 150 0 0 0 0
Outflow ( overland) 0 0 1159 94 1112 90
Outflow (groundwater) 1310 106 1418 115 1441 117
64
56
model. Also, the model is based on an incorrect understanding of the geology and
hydrogeology.
3.19 Before discussing these aspects of the Near Field model, the
following observations can immediately be made, based on a simple review of
DHI’s Table 1, which show several issues of concern with the modelling:
(a) While Table 1 provides the model results of surface water and
groundwater outflows on which Bolivia’s arguments are based, it
also presents information on the water balance of the Silala Near
Field model.174 The law of conservation of mass requires that the
water balance must be closed for the system to be modelled,
meaning that the inflow to the Near Field catchment and model
must equal the total outflow, plus any increase in storage. In this
case, the model has been run as a steady-state simulation,175 for
which inflows and outflows are constant (time-invariant) and must
therefore be equal; there should be no change in storage.176
However, Table 1 defines a change in storage for each scenario.
This is a first indication that the model is not reliable.
(b) In the present case, the inflow to the model is the recharge from
precipitation within the larger groundwater catchment area that
feeds the springs in Bolivian territory. The outflow is the sum of
surface and groundwater cross-boundary flows plus direct loss by
evapotranspiration. As can be seen in Table 1, in the Baseline
Scenario, the inflow is 253 l/s whereas the total outflow, including
174 The water balance over a given time period can be expressed as the equation (P - E = R + ΔS),
in which P stands for precipitation, E for evapotranspiration, R for discharge and S for change in
storage, see CM, Vol. 1, p. 160.
175 DHI Report. BCM, Vol. 5, p. 13.
176 Wheater and Peach (2019), p. 30.
65
57
evapotranspiration, is 266 l/s. Even allowing for the small change
in storage, this is clearly wrong. Similarly, in the Restored
Wetlands Scenario, the inflow is 216 l/s and the total outflow is
220 l/s, which is also incorrect. Only in the No Canal Scenario are
inflow and total outflow equal (221 l/s). This is a further indication
that the scenarios resulting from the DHI Near Field model are not
in accordance with the law of conservation of mass and therefore
render the model unreliable.
(c) The catchment area and recharge from precipitation remain the
same in each of the three scenarios, and therefore the inflow in
each scenario should also remain constant. However, as already
noted above, the inflow to the model is different for each scenario
considered.
In the Baseline Scenario the inflow is 253 l/s; in the No Canal
Scenario the inflow is 221 l/s; and in the Restored Wetlands
Scenario the inflow is 216 l/s.
This means that the difference in the inflows between the Baseline
Scenario and the Restored Wetlands Scenario (i.e. 253 l/s less 216
l/s), equals 37 l/s. And if the inflows are corrected so that they
equal the outflows, as required by conservation of mass,177 the
required inflow to the Baseline scenario is 266 l/s and to the
Restored Wetlands Scenario is 220 l/s, i.e. a difference of 46 l/s.
177 Neglecting the change in storage in each Scenario, which is small and moreover not allowed in
a steady-state simulation.
66
58
This closely approximates the net difference in outflows between
those same two scenarios, of 49 l/s (i.e. a loss of 60 l/s in surface
flows, less a gain of 11 l/s in groundwater flows into Chile).178
The noted difference in outflow is therefore primarily driven by the
difference in inflow, without any apparent change in the catchment
area or recharge to justify such variation.
(d) This difference in inflows raises the question, where would the
“extra” recharge water in the Baseline Scenario go to once the
channels have been removed? The topography and the geology
determine that the recharge must flow to Chile, if not as surface
water, then as groundwater.179
3.20 These initial observations all point to fundamental difficulties with
the Silala Near Field model, in particular, the fact that the recharge from the
Silala groundwater catchment changes for the different scenarios, while the
recharge area and its precipitation remain the same. As Chile’s experts explain,
these changes in inflow are caused in very large part by the way the numerical
model is set up, not by the channelization. Thus, the 30-40% “artificiallyenhanced
flow” alleged by Bolivia is also a result of the modelling exercise, not
of the channels in Bolivia.
178 Wheater and Peach (2019), p. 17.
179 Wheater and Peach (2019), pp. 3, 4 and 8.
67
59
2. Bolivia’s estimation is based on a fundamentally flawed numerical model,
resulting in a gross overestimate of the impact of the wetland
channelization on surface flow rates, by a factor of about 20
3.21 A crucial step in any modelling exercise is the definition of the area
that will be modelled and the conditions at its boundaries (the boundary
conditions). One of the outstanding characteristics of the Near Field model is that
it covers a very small area, equivalent to only 1.1% of the entire Silala
hydrological catchment. The Silala Near Field boundary is drawn around the two
wetlands in Bolivia and the Silala ravine in Bolivia, before crossing into Chile, as
can be seen on Figure 1, but excludes 98.9% of the groundwater catchment area.
3.22 DHI uses a “fixed head” boundary condition at the outer model
(upslope) boundaries in Bolivia. A “fixed head” boundary specifies that the water
table will remain constant, but that means that flows across the boundary can
change. This type of boundary condition is often used where a modelled area is
next to a large lake or hydraulically connected to the sea, and in consequence, the
model can draw for its inflows upon an infinite amount of water. However, as is
obvious, there is no infinite amount of water available in the highlands of the
Atacama Desert – in reality, the inflows are constrained by the available recharge
from precipitation.
68
60
3.23 The effect of a “fixed head” boundary, as DHI itself explains, is
that “flow into the model area may change if the groundwater table changes, e.g.
due to changes in the surface water system”.180 This is so because of Darcy’s law,
one of the basic laws of groundwater flow, which states that “groundwater flow
rate is proportional to the gradient of groundwater potential energy, or head”.181
3.24 The effect of Darcy’s law can be directly observed in the scenarios
of the Near Field model. According to DHI, the channels in the Cajones wetland
are generally less than 50 cm deep and the water tables in both wetlands are
between 15 and 45 cm below surface.182 By removing the channels in the No
Canal Scenario, the groundwater table is elevated accordingly, by a maximum of
50 cm. As a result, the hydraulic gradient between the “fixed head” at the Near
Field boundary and the groundwater table in the wetlands of the No Canal
Scenario is reduced, causing less water to enter into the model area. This explains
why the outflow decreases in the No Canal and Restored Wetlands Scenarios, and
also why the inflow across the “fixed head” boundary is reduced.
3.25 This effect on the inflow is much exaggerated by DHI’s choice of a
“fixed head” boundary for a very small-scale model, i.e. the Near Field model. It
can readily be appreciated that the effect on the hydraulic gradient of a 50 cm
difference in groundwater table height, although relatively small, is
proportionally much more significant when the “fixed head” is set at a distance of
360 m, as in the Near Field model,183 than when the “fixed head” is set at a
distance of 10.500 m, which is the correct boundary of the Far Field, i.e. the
180 DHI Report. BCM, Vol. 5, p. 18.
181 Wheater and Peach (2019), p. 18.
182 DHI Report. BCM, Vol. 3, pp. 12-13, Figures 6 and 7. See Wheater and Peach (2019), p. 18.
183 These are typical distances for the Near Field model, identified by Chile’s experts. See
Wheater and Peach (2019), Fig. 3 a).
69
61
Silala groundwater catchment boundary.184 Chile’s experts show, using DHI’s
simulated groundwater heads, that the change in average hydraulic gradient
differs by a factor of 29, solely due to the geometry.185
184 Wheater and Peach (2019), pp. 19-22.
185 Wheater and Peach (2019), p. 25.
70
62 300 600
Meters
Mercator Projection, WGS 84
g 4380
C:
..0 ,
">' (1J
uj
4370 0
4360 O so
900
BOLIVIA
Area
Enlarged
Change of groundwater gradient using Bolivia's near field boundary
100 150 200
Distance (m)
250 300 350 400
71
63
Figure 2. (a) A typical groundwater head gradient from the near field model boundary to
the wetland; (b) A typical groundwater head gradient from the far field model boundary
to the wetland (Wheater and Peach (2019), p. 23, Figures 3 (a) and (b)).
63
Figure 2. (a) A typical groundwater head gradient from the near field model boundary to
the wetland; (b) A typical groundwater head gradient from the far field model boundary
to the wetland (Wheater and Peach (2019), p. 23, Figures 3 (a) and (b)).
lnacaliri Police
Stati~
CODELCO Intake
]:
C:
0
·;:;
">'
~
LU
4300 0
Kilometers
Mercator Proje<tion,WGS84
1000 2000
\.
Change of groundwater gradient using Bolivia's far field boundary
3000 4000 5000 6000 7000 8000 9000 10000 11000
Distance (m)
72
64
3.26 An elementary calculation, which considers DHI’s estimated
recharge as well as the topographic difference in height between the Near Field
boundary and the Far Field boundary and other parameters, still results in an
exaggeration of the effect of the removal of the channels on the water table
inflow, by a factor of 12.186
3.27 The choice of a Near Field “fixed head” boundary has a similar
exaggerating effect on the factor of “hydraulic resistance”, introduced by DHI in
the Restored Wetlands Scenario. In this scenario, DHI assumes that a peat layer
of up to 60 cm will develop, over long time scales of centuries or more, where the
channels used to be.187 Because peat has relatively low permeability, DHI argues
that this would create a zone of higher hydraulic resistance to groundwater
emerging in the wetlands (“buffer zone”).188 In reality, should this peat layer
indeed be a controlling factor,189 the resistance to flow would cause the
groundwater elevations to rise upslope. However, in the model, the “fixed head”
condition at the model upslope boundary prevents this rise. Thus, the flow across
the boundary into the model area decreases, to compensate for this effect.190
Chile’s experts use a simple calculation to demonstrate that the “buffer zone” has
a disproportionate effect on the inflow in the Near Field area, as compared to the
Far Field area. In their example, the effect is exaggerated by a factor of 23.191
3.28 Similarly, Chile’s experts demonstrate that the combined effect of a
50 cm water table rise in the wetlands and the incorporation of a “buffer zone” of
186 Wheater and Peach (2019), p. 25.
187 DHI Report. BCM, Vol. 5, p. 70.
188 Ibid. Chile’s experts have labelled this “buffer zone”, see Wheater and Peach (2019), p. 27.
189 Chile’s experts note that DHI’s representation of the hydrogeological situation is simplified and
potentially misleading, see Wheater and Peach (2019), p. 27.
190 Wheater and Peach (2019), p. 27.
191 Wheater and Peach (2019), p. 27.
73
65
1 m adjacent to the channels, together with the “fixed head” upslope boundary
condition of the Near Field model, is a decreased inflow in the Near Field model
area by 24%.192 By contrast, those same boundary conditions in a Far Field
model comprising the entire Silala catchment (which is the correct boundary),
have a combined effect of reducing the inflow by only 1.2%.193 This means that
the effect of removing the channels and restoring the wetlands on the inflow (and
hence, on the outflow) is exaggerated by a factor of 20.194 This analysis, based on
simple text book calculations, shows that Bolivia’s exaggeration is directly
attributable to the configuration of the Near Field model, with its “fixed head”
upslope boundary condition at close distance to the wetlands. This forces the
hydraulic gradient to decrease and less water to flow into the model area, exactly
as predicted by Darcy’s law.
3. The DHI Near Field model is built on an incorrect interpretation of the
geology and hydrogeology
3.29 Chile’s and Bolivia’s experts have fundamentally different
interpretations of the geology and hydrogeology of the Silala River basin. The
geological sequences and dates proposed by Bolivia are not supported by recent
geological mapping and radiometric dating, drilling evidence and pumping results
as presented by Chile.195 There is also no evidence for the high permeability fault
along the course of the Silala River, introduced by DHI as an important feature in
its Near Field model.196 On the other hand, the faulting that has been identified by
192 Wheater and Peach (2019), p. 28.
193 Wheater and Peach (2019), p. 28.
194 Wheater and Peach (2019), p. 28.
195 Peach and Wheater (2019), pp. 17-21 and 52.
196 Peach and Wheater (2019), pp. 29-30.
74
66
Chile’s experts at the downstream end of the Silala catchment is not considered in
the DHI model.197
3.30 A numerical model that is built on incorrect geology will not
correctly represent the distribution of areas of high and low permeability in the
basin. It assumes groundwater flow paths and distribution of high and low
hydraulic conductivity that are not supported by evidence and may very well be
different. This means that its predictions have no scientific basis and are highly
likely to be incorrect.198
3.31 In addition, despite acknowledging the differences in origin and
recharge area between the two sets of springs in Bolivia and the existence of
separate aquifer systems,199 DHI does not consider these features in the Near
Field model.200 The groundwaters emerging in the springs are very likely to have
different residence times, due to their different groundwater flow paths coming
from different recharge areas. This is likely to affect their response to the
different scenarios run by the Near Field model, i.e. with or without
channelization and with or without the hypothetical added layer of peat.201
Ignoring this important feature and the complexity of the Silala groundwater
system makes it unlikely that the Near Field model could successfully predict the
behavior of the springs.202
3.32 According to Chile’s experts, the lack of recognition of key
characteristics of the hydrogeology of the Silala River, in particular the ages and
197 Peach and Wheater (2019), pp. 32, 35 and 52.
198 Peach and Wheater (2019), p. 35.
199 DHI Report. BCM, Vol. 4, p. 103.
200 Peach and Wheater (2019), p. 47.
201 Peach and Wheater (2019), p. 47.
202 Peach and Wheater (2019), p. 47.
75
67
sequences of permeable ignimbrites and the existence of separate aquifer and
recharge systems, are serious flaws of the DHI Near Field model and make it
highly improbable that DHI’s scenario predictions could be correct.203
4. Any reduction of the cross-boundary surface flow would anyway be
compensated by an increase of cross-boundary groundwater flow
3.33 DHI recognizes that reduction of surface flows in the No Canal and
Restored Wetlands Scenarios will be compensated by increased groundwater
flows.204 They also acknowledge that the flow direction of the groundwater, as of
the surface water, is from Bolivia towards Chile.205 This means that all the water
in the Silala catchment, less direct loss to evaporation, will ultimately reach
Chile, whether as surface or groundwater flow.206
3.34 The combined outflows in the No Canal and Restored Wetlands
Scenarios are 209 l/s and 207 l/s respectively, amounting to a total reduction of
cross-boundary (surface and groundwater) flows as compared to the Baseline
Scenario (256 l/s), of 18-19%. Taking into account that a simple analysis shows
that the DHI Near Field modelling exaggerates the effect on outflows by a factor
of 20, this indicates that the total reduction of cross-boundary (surface and
groundwater) flows by removing the channels is likely to be a few percent at
most, and therefore, negligible.
203 Peach and Wheater (2019), p. 49 and Wheater and Peach (2019), pp. 29-30.
204 DHI Report. BCM, Vol. 2, p. 266: “11. Without the canals, more water crosses the border as
groundwater.”
205 Ibid.: “5. The observed groundwater levels in the many boreholes established in the Silala
‘Near Field’ and above show a clear flow direction of the groundwater from East to West.”
206 Wheater and Peach (2019), pp. 3, 4 and 8.
76
68
5. The conclusions of the Ramsar Report on wetland degradation at the
Silala are unwarranted and are contradicted by recent evidence provided
by DHI and other expert reports
3.35 In addition to the DHI expert report, Bolivia has presented a report
of the Ramsar Advisory Mission N° 84 on the Los Lípez Ramsar Site (Ramsar
Report).207 The purpose of this report is to evidence significant degradation of the
Silala wetlands in Bolivia.
3.36 The Ramsar Mission was requested by Bolivia in July 2016,208
shortly after Chile had lodged its current Application, on 6 June 2016. The
Ramsar Report is based on information provided by Bolivia and one site visit to
the Los Lípez Ramsar Site, in November 2016.209 It contains several statements
that appear uncritically to reflect Bolivia’s position on the Silala and that are not
supported by the relevant evidence, including the DHI expert report and other
recent studies undertaken by Bolivia.
3.37 The Ramsar Mission classifies the Silala groundwater system as a
“non-renewable aquifer on the geological scale.”210 Conclusion 9 of the Ramsar
Report states: “Studies with stable isotopes have shown that the waters that
emerge in the Silala springs are fossil waters dating back to more than
207 Ramsar Report. BCM, Vol. 5, Annex 18. The Silala wetlands are located at the north-western
boundary of the Los Lípez Ramsar Site in Bolivia. Chile notes that the Ramsar Report includes a
discussion of several lagoons within the Los Lípez Ramsar site, some quite far removed from the
Silala and none of which bear relation to the Silala wetlands.
208 Note N° VRE-Cs-58/2016 from the Ministry of Foreign Affairs of Bolivia to the Senior
Advisor for the Americas of the Ramsar Convention Secretariat, 27 July 2016. CR, Vol. 2,
Annex 97.
209 Ramsar Report. BCM, Vol. 5, p. 101.
210 Ramsar Report. BCM, Vol. 5, p. 149.
77
69
10,000 years. In other words, these are waters that are not renewed by natural
recharges of meteoric waters in the local aquifer.”211
3.38 These statements are contradicted by DHI, who unequivocally
confirm that the waters of the Silala are largely from recharge:
“The hydrological catchment, Catchment B, can sustain a flow of
151-374 l/s from recharged water which is in the same order of
magnitude as the observed surface water (160-210 l/s) and
estimated cross border groundwater flow in the order of
(100-230 l/s) (Annex F and Annex H).
Overall, the analysis indicates that a large proportion of the water
feeding the wetland is from recharge from rainfall and snow melt in
the hydrological catchment.”212
3.39 The Ramsar Mission also uncritically reproduces Bolivia’s position
that the channelization affected the extension of the wetlands, as follows:
“The wetlands found in the Silala area have been highly affected by
the construction of the water-catchment canals started in 1908. At
present, there are only vestiges of the original wetlands that used to
cover an area of about 141,200 m2, or 14.1 hectares. The current
surface area of the wetlands covers only about 6,000 m2, or 0.6 ha,
which are surrounded by the water catchment works and artificial
canals (SERGEOMIN, 2003).”213
3.40 The SERGEOMIN (2003) report, the only and direct source for this
statement, does not make any reference to historical studies or scientific
investigations to confirm its estimation at 14.1 ha of the area originally covered
by the wetlands. Nor does it provide any evidence for the alleged reduction of
211 Ramsar Report. BCM, Vol. 5, p. 167.
212 DHI Report. BCM, Vol. 2, p. 290.
213 Ramsar Report. BCM, Vol. 5, p. 163. Chile notes that the year 1908 coincides with the Bolivian
concession, but not with the year the channelization was undertaken, which was in 1928. The
reproduced text of the Ramsar Report is taken from Bolivian Geology and Mining Survey
(SERGEOMIN), Study on Hydrographic Basins, Silala Springs Basin, Basin 20, June 2003
(henceforth “SERGEOMIN 2003”), pp. 59-60. CR, Vol. 2, Annex 94.
78
70
this surface area to a mere 0.6 ha in 2003. Indeed, elsewhere in the same
SERGEOMIN (2003) report, the total surface of the wetlands in 2003 is
estimated much higher, at 108.600 m2 or 10.8 ha.214 The very low estimate of the
Silala wetland extension relied on by the Ramsar Mission is also contradicted by
Bolivia’s own more recent studies, including a 2016 study of the Bolivian
Ministry of Environment and Water, which estimates the total extension of the
wetlands at the Silala at 10.89 ha in June 1986, and at 9.81 ha in June 2010.215
3.41 A more complete study of the wetlands extension was
commissioned by Bolivia in 2017 (the Castel study). It consists of a multitemporal
analysis using satellite imaging in order to assess changes at the
Orientales and Cajones wetlands, between 1975 and 2017.216 The Castel study
estimates the current wetlands as fluctuating between 8.01 and 6.21 ha during the
wet season, and between 6.75 and 2.16 ha during the dry season (1975-2000); and
between 5.88 and 3.58 hectares during the wet season, and between 3.65 and
1.92 ha during the dry season (2002-2017).217 Its estimates again do not come
close to the 0.6 ha relied on by the Ramsar Mission from the SERGEOMIN
(2003) report. Castel refrains from giving any estimates of the wetland extension
prior to the channelization, for reason that no satellite images are available from
the early twentieth century that could support such estimation.218
214 SERGEOMIN (2003), pp. 26 and 65. CR, Vol. 2, Annex 94.
215 Ministry of the Environment and Water of Bolivia, Characterization of Water Resources in the
Southwest of the Department of Potosí – Municipality of San Pablo de Lipez “Wetlands of Silala
Valley and Adjacent Sectors” (Volume II), July 2016, p. 40. CR, Vol. 2, Annex 96. The Ramsar
Report cites the 2016 study, noting a reduction by 1.08 ha in the wetlands surface area between
1986 and 2010, apparently without noticing that the estimates of the Ministry do not coincide with
the very low estimate relied upon by the Ramsar Mission from the SERGEOMIN 2003 report. See
Ramsar Report, BCM, Vol. 5, p. 163.
216 Castel, p. 4. CR, Vol. 2, Annex 98.
217 Castel, p. 38. CR, Vol. 2, Annex 98.
218 Castel, p. 38: “[T]he Silala Spring high altitude wetlands have remained in this state of
intervention since the beginning of the XX century, therefore they could not be analyzed through
79
71
3.42 The Castel study also provides no evidence for the “progressive
degradation” of the wetlands which the Ramsar Mission claimed to have
observed during its November 2016 site visit.219 The Castel study affirms that for
the entire period studied (1975-2017):
“No significant long-term changes were noted during both periods
studied in the surface of the high altitude wetlands. Both the
Landsat images and the high resolution images show that although
there is significant seasonal variability, there is no trend towards a
decrease in the total surface area of the high altitude wetlands.”220
3.43 Castel’s study reinforces the results presented in Chile’s Memorial
that show strong seasonal and inter-annual variability of the wetlands, rather than
long term change.221 Recent results from Chile’s science team show that the
current extents of the Cajones and Orientales wetlands wholly fill the available
valley floor and seasonally expand up the adjacent hillslopes, in the same way as
a wholly undisturbed wetland in Chile.222 So claims by the Ramsar Mission on
wetland degradation appear wholly unfounded and counterfactual.
3.44 But even if it were the case that the Orientales and Cajones
wetlands suffered from the 1928 channelization, this is not attributable to Chile.
Bolivia authorized the FCAB works in its territory and could have restored the
wetlands many years ago.
satellite imaging over a period of time in a natural state, without intervention.” CR, Vol. 2,
Annex 98.
219 Ramsar Report. BCM, Vol. 5, p. 163. Chile notes that the November 2016 site visit occurred at
the end of the dry season.
220 Castel, p. 38. See also pp. 15, 21, 25, 28 and 30. CR, Vol. 2, Annex 98.
221 Alcayaga, H., Characterization of the Drainage Patterns and River Network of the Silala River
and Preliminary Assessment of Vegetation Dynamics Using Remote Sensing, 2017, p. 31, Fig. 16.
CM, Vol. 4, Annex I.
222 Wheater and Peach (2019), p. 44.
80
72
C. Conclusion: The impact of the 1928 channelization due to reduced
loss to evapotranspiration, is limited to no more than 2% of the current
cross-boundary surface flow; any additional impact argued by Bolivia is
grossly exaggerated
3.45 As explained, the relevant remaining disagreement between the
Parties concerns the quantitative effect of the 1928 channelization in Bolivia on
the cross-boundary surface flow which, according to Bolivia’s expert DHI, would
be 30-40% less without the channelization.223 According to Chile’s experts, the
effect of the channelization is minimal, while the very large estimate by DHI is
implausible and defies common sense.224 This disagreement is significant for
Bolivia’s case, because its position concerning “artificially-enhanced flows” is
dependent upon its view that the science indicates that the channelization results
in additional flows that enhance the natural flow by a factor of 30-40%.
3.46 Chile’s experts have demonstrated that the DHI Near Field
modelling has important flaws that have led to the exaggerated and incorrect
results relied on by Bolivia as a base for its second and third Counter-Claims.
DHI’s estimates are based on a numerical model of only a small part of the Silala
River basin, called the Near Field model, built on an incorrect understanding of
the geology and hydrogeology of the Silala River system and using incorrectly
defined boundary conditions. DHI’s projections of long-term peat growth are also
highly speculative.
3.47 Both Parties’ experts agree that the 1928 channelization may have
had a minor impact on the cross-border surface flow due to reduced
evapotranspiration in the Bolivian wetlands, estimated by both at no more than
223 DHI Report. BCM, Vol. 2, p. 266.
224 Wheater and Peach (2019), p. 2.
81
73
2% of the current cross-boundary flow. Indeed, Chile’s recent estimates of
evaporation from satellite data show no detectable effect of channelization.225
3.48 It follows that there is no factual or scientific foundation for
Bolivia’s claim that the channels in Bolivia have resulted in a 30-40%
“artificially-enhanced flow”. Moreover, whatever the impact of the
channelization – it is marginal – the distinction between natural and “artificial”
flow is untenable in international law as was demonstrated in Chapter 2.
225 Wheater and Peach (2019), p. 45.

83
75
SUBMISSIONS
With respect to the counter-claims presented by the Plurinational State of Bolivia,
Chile requests the Court to adjudge and declare that:
(a) The Court lacks jurisdiction over Bolivia’s Counter-Claim a),
alternatively, Bolivia’s Counter-Claim a) is moot, or is otherwise
rejected;
(b) Bolivia’s Counter-Claims b) and c) are rejected.
Ximena Fuentes T.
Agent of the Republic of Chile
15 February 2019

85
Expert Report
Wheater, H.S. and Peach, D.W., Impacts of Channelization
of the Silala River in Bolivia on the Hydrology of the
Silala River Basin

87
IMPACTS OF CHANNELIZATION OF THE SILALA RIVER IN
BOLIVIA ON THE HYDROLOGY OF THE SILALA RIVER BASIN
Drs. Howard Wheater and Denis Peach
January 2019

89
iii
ABOUT THE AUTHORS
Dr. Howard Wheater
Dr. Howard Wheater is the Canada Excellence Research Chair Laureate in Water
Security at the University of Saskatchewan, where he was founding Director of
the Global Institute for Water Security, and Distinguished Research Fellow and
Emeritus Professor of Hydrology at Imperial College London. A leading expert in
hydrological science and modelling, he has published more than 200 refereed
articles and 6 books. He is a Fellow of the Royal Society of Canada, the UK’s
Royal Academy of Engineering and the American Geophysical Union. He was
awarded the 2018 International Prize for Water (Dooge Medal) by UNESCO, the
World Meteorological Organisation and the International Association for
Hydrological Sciences and the 2006 Prince Sultan bin Abdulaziz International
Prize for Water. He has initiated and led national and international research
programmes in the UK and Canada, and has advised states, provinces and national
governments on flood, water resource and water quality issues. He was instructed
by Hungary and Argentina at the International Court of Justice, and recently sat
on the Court of Arbitration concerning the Indus Waters Treaty. He was, until
2014, vice-chair of the World Climate Research Programme’s Global Energy and
Water Cycle Exchange (GEWEX) project and leads UNESCO’s GWADI arid
zone water program. In Canada, he led the Changing Cold Regions Network,
focused on the analysis and prediction of hydrological change in western Canada,
and the Global Water Futures Program, focused on managing water futures in
Canada and other cold regions where global warming is changing landscapes,
ecosystems, and the water environment. His role as Chair of the Council of
Canadian Academies Expert Panel on Sustainable Management of Water in the
Agricultural Landscapes of Canada saw release of a report in February 2013
entitled Water and Agriculture in Canada: Towards Sustainable Management of
90
iv
Water Resources. In 2018 he was the only non-US member of a US National
Academies panel that reported on Future Water Priorities for the Nation.
Dr. Denis Peach
Dr. Denis Peach spent nine years as the Manager of the British Geological Survey
(BGS) Groundwater Programme, then 6 years as BGS Chief Scientist. He is a
hydrogeologist with broad scientific interests and 44 years of experience which
includes work for a UK water authority, overseas work in tropical
hydrogeological environments, small island hydrogeology and work for
international consultants in arid zone hydrogeology. His particular scientific
interests include groundwater modelling which he developed in BGS, arid zone
hydrogeology, and Chalk and Karst hydrogeology. He has been a Vice President
of the Geological Society of London (GSL), is a Visiting Professor at Imperial
College, London and University of Birmingham and he has given the Ineson
Distinguished lecture at the GSL. He has led numerous national geological and
hydrogeological research programmes in the UK and has sat on many national
research programme boards and national water resource strategic committees. He
currently carries out research with BGS and Imperial College and recently
provided advice to the University of Saskatchewan and UK engineering
consultants.
91
v
TABLE OF CONTENTS
LIST OF FIGURES............................................................................................... vii
LIST OF TABLES ............................................................................................... viii
LIST OF ACRONYMS AND ABBREVIATIONS............................................... ix
1 INTRODUCTION......................................................................................... 1
1.1 Experts’ Terms of Reference .................................................................. 1
1.2 Background to the report ........................................................................ 2
1.3 Structure of the report............................................................................. 2
2 SUMMARY................................................................................................... 3
3 POINTS OF SCIENTIFIC AGREEMENT BETWEEN THE PARTIES’
EXPERTS CONCERNING THE HYDROLOGY OF THE SILALA
RIVER ........................................................................................................... 7
3.1 Agreement that the Silala River is an international watercourse............ 7
3.2 Agreement concerning the likely Groundwater Catchment area............ 9
3.3 Agreement concerning the potential for the effects of
channelization in Bolivia to affect surface water flows at the border .. 10
3.4 Agreement concerning the effects of drainage on wetland
evaporation ........................................................................................... 13
4 POINTS OF SIGNIFICANT SCIENTIFIC DISAGREEMENT
BETWEEN THE PARTIES’ EXPERTS CONCERNING THE
HYDROLOGY OF THE SILALA RIVER................................................. 14
4.1 Effects of wetland drainage on groundwater discharges to the
wetlands ................................................................................................ 14
4.1.1 Water balance considerations ................................................... 15
92
vi
4.1.2 Increased hydraulic gradients ................................................... 18
4.1.3 Reduced hydraulic resistance.................................................... 27
4.1.4 Other modelling issues.............................................................. 29
4.2 Other areas of disagreement ................................................................. 31
5 NATURAL VARIABILITY AND FUNCTIONING OF THE
BOLIVIAN WETLANDS AND TOPOGRAPHIC CONSTRAINTS........ 31
6 CONCLUSIONS ......................................................................................... 43
7 REFERENCES ............................................................................................ 46
APPENDIX 1 ........................................................................................................ 47
93
vii
LIST OF FIGURES
Figure 1. Silala River topographic catchment (outlined in black) and groundwater
catchment (outlined in green).................................................................................. 9
Figure 2. Simulated groundwater potential (head) in the Ignimbrite aquifer,
adapted from a DHI’s figure (BCM, Vol. 3, p. 488)............................................. 19
Figure 3a. A typical groundwater head gradient from the near field model
boundary to the wetland. ....................................................................................... 23
Figure 3b. A typical groundwater head gradient from the far field model boundary
to the wetland. ...................................................................................................... 23
Figure 4. Effect of fixed groundwater head at near-field boundary on spring
discharge................................................................................................................ 26
Figure 5. Effect of fixed groundwater head at a more realistic 10.5 km boundary
on spring discharge................................................................................................ 26
Figure 6. Effect of low conductivity zone on spring discharge – near field fixed
head. ...................................................................................................................... 28
Figure 7. Effect of low conductivity zone on spring discharge – far field fixed
head. ...................................................................................................................... 29
Figure 8. Location of the Quebrada Negra wetland within the Silala River
topographic catchment in Chile (Muñoz and Suárez, 2019)................................. 33
Figure 9. Photograph of the Quebrada Negra wetland, taken from the northern
slope (Muñoz and Suárez, 2019)........................................................................... 34
Figure 10. Photograph taken at the Quebrada Negra wetland, looking upstream
(Muñoz and Suárez, 2019). ................................................................................... 34
Figure 11. Photograph of the Quebrada Negra wetland, taken from the southern
slope (Muñoz and Suárez, 2019)........................................................................... 35
Figure 12. Photograph of Bolivian wetland (BCM, Vol. 2, p. 333)...................... 35
Figure 13. Quebrada Negra, Cajones and Orientales wetlands average NDVI
distribution from July to November 2018 (Muñoz and Suárez, 2019). ................ 37
Figure 14. Cross section of vegetation cover (NDVI>0.2) and topography of the
Quebrada Negra wetland (Muñoz and Suárez, 2019). .......................................... 38
94
viii
Figure 15. Cross section of vegetation cover (NDVI>0.2) and topography of the
Cajones wetland (Muñoz and Suárez, 2019)......................................................... 39
Figure 16. Cross section of vegetation cover (NDVI>0.2) and topography of the
Orientales wetland (Muñoz and Suárez, 2019). .................................................... 39
Figure 17. Layout of the monitoring wells in the Quebrada Negra wetland (Muñoz
and Suárez, 2019).................................................................................................. 42
Figure 18. Contour lines of groundwater levels (m.a.s.l.), at shallow piezometers,
measured during September 2018 at the main grassland of the Quebrada Negra
wetland (Muñoz and Suárez, 2019)....................................................................... 43
LIST OF TABLES
Table 1. DHI’s results of its modelling of different scenarios (reproduced from
BCM, Vol. 5, p. 67, Table 1)................................................................................. 17
Table 2. Area covered by vegetation in the Quebrada Negra, Cajones and
Orientales wetlands, from July to November, 2018 (Muñoz and Suárez, 2019). . 40
Table 3. Annual ETa,NDVI, Mean and Standard Deviation (S.D.) in mm/year
estimated for the Quebrada Negra, Cajones and Orientales wetlands (after Muñoz
and Suárez, 2019).................................................................................................. 41
95
ix
LIST OF ACRONYMS AND ABBREVIATIONS
BCM – Counter-Memorial of the Plurinational State of Bolivia
CM – Memorial of the Republic of Chile
DHI – Danish Hydraulic Institute
ETa,NDVI – Actual evaporation rate estimated using NDVI data
ha – hectares
km2 – square kilometres
l/s – litres per second
m – metres
m/day – metres per day
m/s – metres per second
NDVI – Normalized Difference Vegetation Index

97
1
1 INTRODUCTION
1.1 Experts’ Terms of Reference
In the context of the dispute between the Republic of Chile and the Plurinational
State of Bolivia concerning the status and use of the waters of the Silala, to be
heard before the International Court of Justice, the Republic of Chile has
requested our independent expert opinion, as follows:
“Questions for Dr. Howard Wheater, as a hydrological engineer:
(i) What are the major points of scientific agreement between
the Experts of Bolivia and those of Chile concerning the
hydrology of the Silala River?
(ii) What are the major points of scientific disagreement between
the Experts of Bolivia and those of Chile concerning the
hydrology of the Silala River?
(iii) What new evidence has been produced, since Chile
submitted its Memorial in July 2017, concerning the effect of
the channelization of the flow on Bolivian territory on the
watercourse of the Silala River that flows from Bolivia into
Chile?
Questions for Dr. Denis Peach, as a hydrogeologist:
(i) What new evidence has been produced, since Chile
submitted its Memorial in July 2017, concerning the
understanding of the geology and hydrogeology of the Silala
River?
(ii) Does the hydrogeological conceptual understanding and
parameterisation of the numerical models of Bolivia’s
Expert, the Danish Hydraulic Institute (DHI), provide an
adequate basis to quantify the effects of channelization on the
surface water and groundwater flows from Bolivia to Chile?
(iii) Could the flow from groundwater-fed springs in the Cajones
and Orientales springs have been significantly enhanced by
the use of explosives?”
98
2
In this joint report we address the three questions to Wheater. A separate report
(Peach and Wheater, 2019) addresses the questions to Peach.
1.2 Background to the report
This report follows two expert reports, Wheater and Peach (2017) and Peach and
Wheater (2017), which were requested by the Republic of Chile as a contribution
to its Memorial to the International Court of Justice. At the time, the core of the
dispute between Chile and Bolivia was whether or not the Silala River is an
international watercourse.
Following submission of Bolivia’s Counter-Memorial (BCM), and in particular
the report of Bolivia’s consultants, the Danish Hydraulic Institute (DHI), on 3
September 2018, we now understand that there is agreement between the parties
on the central point that the Silala River naturally flows from Bolivia to Chile and
is an international watercourse. As we show below, there is also general
agreement between Bolivia’s experts and ourselves about the nature and
functioning of the natural hydrological system.
The core of the dispute between Chile and Bolivia is now the quantitative effect of
the channelization of the Silala, on Bolivian territory, on the cross-boundary flow.
DHI estimates that the natural flows without drainage and channelization would
be 30-40% less than the current situation (BCM, Vol. 2, p. 266). We disagree. In
our opinion the very large estimates made by DHI are implausible, and indeed
defy common sense.
1.3 Structure of the report
In section 2, we summarize our conclusions. We set out the points of agreement
between the parties’ experts in section 3, and then in section 4 explain the points
99
3
of significant disagreement and the causes thereof. In section 5 we introduce new
information from an undisturbed wetland in the Silala basin in Chile that is
comparable to the Cajones and Orientales wetlands in Bolivia, and hence
comment on the hydrological functioning of Bolivia’s Cajones and Orientales
wetlands. Section 6 presents our conclusions.
While this co-authored report represents our joint opinion, Wheater is the lead
author of the report.
2 SUMMARY
We are pleased to note important areas of agreement between Bolivia’s experts,
the Danish Hydraulic Institute (DHI), and ourselves.
We agree in general terms about the hydrology of the Silala River and its
catchment area. A central point is that the Silala River, which rises in two sets of
springs in Bolivia that support the Cajones and Orientales wetlands, flows
naturally from Bolivia to Chile and is an international watercourse. The river is
primarily fed by groundwater, and interacts with groundwater along its course.
The groundwater is recharged from an extended groundwater catchment, termed
by DHI the hydrological catchment. In addition to the surface water flow in the
river from Chile, there is an extensive groundwater system also flowing from
Bolivia to Chile, recharged from the groundwater catchment area, and possibly
further afield. The recharge from the groundwater catchment area, apart from the
water lost in evaporation in the wetlands of the basin, will flow from Bolivia to
Chile, either as surface water or as groundwater.
We also agree with DHI that the channelization that occurred in Bolivia in the
1920s, through excavation of channels in the Bolivian wetlands and lining the
main river channel downstream of the wetlands, will have affected the river flows.
We agree that the channels in the wetlands may have reduced evaporation losses,
100
4
and we agree that this effect will be small, no more than 2% of the current flow to
the border (in our current view, based on new Chilean data, probably much less).
We further agree that there may have been some changes to river-groundwater
interactions downstream of the wetlands due to the main river channelization, but
these affects also will be small.
There is however one major point of disagreement. DHI suggest that groundwater
inflows to the Cajones and Orientales springs may have been affected by the
channelization due to a change in the gradient of groundwater flow, and by the
removal of peat overlying the springs. They claim these latter effects are large, so
that the total impacts of the channelization account for 30-40% change in surface
water flows. We agree that these effects may occur, but find DHI’s large estimates
to be implausible. These estimates are wholly based on hydrological modelling of
a small area around the springs (the Near Field), which we find to be
fundamentally flawed.
Errors in the modelling include the fact that the underlying geology is misrepresented,
and the boundary conditions for the model are inappropriate. In
particular, water table conditions at the model upslope boundary are held constant.
One effect of this is that the inflows to the model change significantly for the
different scenarios investigated by DHI, whereas in reality, the recharge arises
from the precipitation over the groundwater catchment, and is unaffected by the
channelization. And because the inflows to the model change, the model outputs
change too. The combined surface water and groundwater flows in DHI’s model
change by 18-19% for the different scenarios. In reality of course, this recharge
can only flow to Chile – either as surface water or as groundwater.
We demonstrate below, using simple calculations, that this erroneous boundary
assumption can exaggerate the effects of water table rise and peat cover by a
factor of 20, and appear to explain DHI’s exaggerated estimates. In our opinion,
the effects of water table rise and peat cover will be minor, a few percentage at
101
5
most of the cross-border flow. And any reduction in surface water flow would be
accompanied by a corresponding increase in groundwater flow, both flowing
down the topographic and hydraulic gradient to Chile.
We address in summary, the three questions posed to us by Chile. Further detail is
provided in the full report that follows:
(i) What are the major points of scientific agreement between the Experts of
Bolivia and those of Chile concerning the hydrology of the Silala River?
We and Bolivia’s experts agree that:
1. The Silala River flows naturally from Bolivia to Chile. The river rises in
two sets of springs in Bolivia, which maintain the Cajones and Orientales
wetlands.
2. The river is primarily fed by groundwater and interacts with groundwater
along its course to the border and beyond.
3. In addition, there are substantial groundwater flows from Bolivia to Chile,
likely of an equivalent magnitude to the surface water flows.
4. Construction of drainage channels and river channelization in the 1920s
will have had some effect on the flow. An increase in flow due to these
works is expected.
5. The impact of drainage on evaporation from the wetlands is small.
(ii) What are the major points of scientific disagreement between the Experts
of Bolivia and those of Chile concerning the hydrology of the Silala River?
We and Bolivia’s experts disagree about the magnitude of the impact of the
drainage works. In our opinion, Bolivian estimates of a 30-40% effect on flows
are implausible. These estimates have been produced by a Near Field model of
102
6
surface water-groundwater interactions. We have shown that the model is based
on incorrect geology, that simple calculations show that incorrect assumptions of
the model’s boundary conditions lead to an overestimate of the impacts, by a
factor of approximately 20, and that the change in inputs to the model is
unrealistic.
(iii) What new evidence has been produced, since Chile submitted its Memorial
in July 2017, concerning the effect of the channelization of the flow on Bolivian
territory on the watercourse of the Silala River that flows from Bolivia into Chile?
New studies based on detailed monitoring of an undisturbed Chilean wetland
within the Silala basin, coupled with high resolution remote sensing data, show
that Bolivian and Chilean wetlands continue to fully occupy the valley floor, and
seasonally extend up the base of adjacent hillslopes. The condition of the wetland
vegetation, as indicated by remote sensing, is similar in all three wetlands, and
associated estimates of actual evaporation suggest that the highest evaporation
rates are observed from the Cajones and Orientales wetlands, some 10% greater
than that of the undisturbed Quebrada Negra wetland. At least from the satellite
data, it appears that there has been no significant change in evaporation associated
with the channelization of the Bolivian wetlands, and hence no effect of
evaporation change on river flows.
In summary, we remain confident that the effects of the drainage works on
evaporation are quite limited, as stated in Chile’s Memorial (CM), at most
equivalent to a flow of 2-3 l/s on average, i.e. some 2% of the natural flow, but in
the light of our recent results, probably less. Other effects will be similarly small.
Bolivia’s estimates of 30-40% changes in river flow are due to errors in DHI’s
modelling and are implausible. We also reiterate that there is no doubt that the
103
7
Silala River is an international watercourse, and we are pleased to note the
agreement of Bolivia’s experts on this point.
3 POINTS OF SCIENTIFIC AGREEMENT BETWEEN THE
PARTIES’ EXPERTS CONCERNING THE HYDROLOGY OF THE
SILALA RIVER
3.1 Agreement that the Silala River is an international watercourse
Our evidence in Chile’s Memorial and the supporting scientific annexes showed
that:
The Silala River rises at two sets of springs (Cajones and Orientales) in Bolivia
and flows along the natural topographic gradient from Bolivia to Chile, crossing
the border in a ravine. The geomorphological history shows that the river has
flowed from Bolivia to Chile in its current ravine for at least 8000 years.
The characteristics of measured river flow at the border show that the dominant
source of the river water is groundwater. In addition to the Cajones and Orientales
springs, we have found other spring flows contributing additional surface water
flows to the river and significant groundwater flows at depth, downstream of the
border. There are also areas where the river flow loses water to the underlying
groundwater system.
This was summarized by Wheater and Peach (CM, Vol. 1, p. 177). We stated that:
• ‘[T]he topography of the Silala River catchment area is such that natural
drainage from […] Bolivia flows across the international border between
Bolivia and Chile.’
• ‘[W]hile the source areas for the perennial flow at the border lie in two
major sets of groundwater springs in Bolivia (the water sources for the
104
8
Orientales and Cajones wetlands), the river interacts with groundwater
throughout its subsequent course.’
• ‘This [the Silala River] is […] “a system of surface waters and
groundwaters constituting by virtue of their physical relationship a unitary
whole and normally flowing into a common terminus.”’
We are pleased to note that Bolivia’s consultants DHI agree with the above
statements. They confirm that the Silala River is ‘a coupled groundwater-surface
water system […] extending across the border’ (BCM, Vol. 2, p. 266). They also
note that numerous additional springs occur downstream of the Orientales
wetlands and add to the river flow (BCM, Vol. 2, pp. 368-369). They argue that
the channelization works carried out in Bolivia in 1928 have influenced the
magnitude of flow at the border, but note that ‘The canals have changed the
amount of discharge from the Silala springs but not the direction of natural
outflow from the Silala wetlands’, ‘in a situation without the canals, the discharge
direction is towards Chile’ and ‘In a situation without the canals, it is not possible
that all surface water discharged from the wetlands infiltrate from the confluence
point to the border’ (BCM, Vol. 2, pp. 266-267). Further, ‘groundwater level
gradients and hydrogeological properties clearly indicate groundwater flow from
Bolivia to Chile’ and ‘the groundwater flow across the border is at least of the
same order of magnitude as surface water discharge at the border’ (BCM, Vol. 5,
p. 84).
Thus, in addition to the agreement between the parties that the Silala River
naturally flows from Bolivia to Chile, there is agreement between Bolivia’s
experts and ourselves about the existence of substantial groundwater flows from
Bolivia to Chile. It is clear that, whether as surface water, or as groundwater, the
water from the Silala River catchment area flows from Bolivia to Chile. There is
also general agreement concerning the nature and functioning of the natural
hydrological system, as we show below.
105
9
3.2 Agreement concerning the likely Groundwater Catchment area
In Chile’s Memorial the topographic catchment of the Silala River was defined
(CM, Vol. 1, p. 140, Figure 2), i.e. the area that drains surface or near-surface
flows naturally to the border, but the possibility of groundwater inflows from
areas beyond that, within Bolivia, was noted (CM, Vol. 4, p. 273, Figure 7-1). Our
current best estimate of the larger area contributing groundwater recharge to the
Silala River, based on topographic and geological analysis, is shown in Figure 1,
below.
Figure 1. Silala River topographic catchment (outlined in black) and
groundwater catchment (outlined in green).
Kilometers
,\krt.itor ~jl1.l.oa,. WU!> 84
Military
O ,Post
"
1~nta/es
~ SILALA RIVER BASIN
GROUNDWATER
CATCHMENT
----SILALA RIVER
TOPOGRAPHIC
CATCHMENT
\_ 1. ~ BRAZIL
PERU ( ~
.· 1 BOLIVIA Area
·· z Enlarged "" / ~ CHILE
0 PARAGUAV
v
'<
i J }
ARGENTINA /_
..,.
106
10
This area is very similar to that identified by DHI as the ‘hydrological catchment’
(BCM, Vol. 2, p. 275, Figure 5) of the Silala River (with minor differences due to
the use of a different Digital Elevation Model), although both we and DHI (BCM,
Vol. 4, p. 103) acknowledge that it is possible that there may be additional
groundwater contributions from further, more distant, sources.
3.3 Agreement concerning the potential for the effects of channelization in
Bolivia to affect surface water flows at the border
The original concessions for use of the waters of the Silala date from 1906 (from
the Government of Chile) and 1908 (from the Government of Bolivia), and at that
time small structures were put in place in the river near the border to allow
diversion of the waters into collector channels and pipes for transmission to
downstream users. Some 20 years later, in 1928, modifications were made to the
upstream channel in Bolivia. As we noted in our previous report, ‘[e]arth channels
of the order of 0.6 x 0.6 m cross-section were constructed and subsequently lined
with stone. They thus act as drains and are able to receive water from the wetland
soils (and to release water back to riparian soils).’ (CM, Vol. 1, p. 134).
We are grateful to Bolivia for providing further details of the geometry of the
drains and recent photographs (BCM, Vol. 1, pp. 41-42, Figures 15 and 16),
including the areas where they have been blocked in recent years to divert inchannel
flow to adjacent wetlands (BCM, Vol. 2, p. 370, Figure 4). Bolivia also
provides data on the effect of the drains on groundwater elevations in the wetland
source areas (BCM, Vol. 3, pp. 12-13, Figures 6 and 7). These show that the
current water table depths in the drained wetlands range from 0.15 to 0.4 m below
surface in the Cajones wetlands (Bolivia’s ‘Northern’ wetland) and from 0.15 to
0.45 m in the Orientales wetland (Bolivia’s ‘Southern’ wetland). In other words,
107
11
instead of having standing water on the surface, water levels have reduced, but by
less than 50 cm.1
Both we and Bolivia’s experts agree that there will be some effect of these
drainage works and the channelization of the river on the flows at the border.
However, while both sides agree on the various effects that are possible, and agree
about the magnitude of some of these, there is strong disagreement about the
magnitude and significance of others. DHI suggests that in total these effects
could give rise to a potential 30-40% change in flows, arguing that channelization
increased river flows, and that long term restoration of the wetland peat soils
could lead to further reductions in flow (BCM, Vol. 2, p. 266). In our view these
estimates are wildly exaggerated and implausible, for reasons that we explain
below.
Concerning changes in evaporation from the wetlands, we noted, for example,
‘[w]hile active, the channel works are likely to have reduced the extent of surface
water in the wetlands and hence reduced the direct loss of water to evaporation
[…]. Any resulting reduction in evaporation would potentially provide additional
water for surface discharge, including cross-border flows.’ (CM, Vol. 1, p. 134).
Bolivia agrees, and the various estimates presented by both sides are discussed in
2.4 below. There is general agreement that while changes to evaporation are
expected, they will have minor effects on river flows.
We agree too that the channelization may have affected the interaction of surface
water and groundwater downstream of the wetlands, in Bolivia, although in our
opinion these effects will be small. For example, DHI states that infiltration from
the river will have been reduced in reaches where the groundwater tables are
lower than terrain level (BCM, Vol. 2, p. 276), thereby increasing the surface
flows downstream. This is certainly possible, although as can be seen from
1 Assuming that the spatially-variable standing water had a depth of less than 5 cm.
108
12
Bolivia’s profiles of water table elevation (BCM, Vol. 2, pp. 285-287, Figures 11-
13), groundwater levels are predominantly higher than channel bed levels along
the main channel and much of the Cajones (northern) and Orientales (southern)
tributaries, so that groundwater would be likely to be contributing flow to the
stream under those conditions. Chile’s observations downstream of the border
gave infiltration losses from a losing reach of 3.3 l/s over an approximately 2 km
reach (CM, Vol. 5, p. 489), whereas we estimate that the length of potentially
losing reaches in Bolivia is 1.4 km, which suggests that any surface water losses
to groundwater in Bolivia are likely to be quite limited. It is also important to note
that, as stated by Bolivia, this is a coupled surface water-groundwater system.
DHI states (BCM, Vol. 2, p. 266) that ‘[t]he observed groundwater levels in the
many boreholes established in the Silala “Near Field” and above show a clear
flow direction of the groundwater from East to West. Together with evidence
from boreholes of a pervious and water holding aquifer this proves the presence of
cross border groundwater flow into Chile.’ Both the Bolivian interpretation of the
hydrogeology and our own agree that water lost from the river to groundwater
will still flow to Chile, albeit as groundwater rather than surface water flow.
There is therefore general agreement between Bolivia’s experts and ourselves that
the 1928 drainage works will have affected surface flows across the border due to
reduced direct loss of water by evaporation and possibly by infiltration, but that
the effects of these on surface and groundwater flow from Bolivia to Chile are
minor.
Bolivia also proposes that major changes have occurred to the groundwater
discharges that feed the Bolivian wetlands, due to changing groundwater levels
associated with the construction of channels in the wetlands, and to the effects of
the wetland peat soils and their possible future long term evolution, creating
hydraulic resistance to groundwater discharge to the wetlands (BCM, Vol. 5,
p. 83). We agree that such effects could occur, but in our opinion these will also
109
13
be very minor. Bolivia’s estimates are infeasible and appear to arise, in the major
part, due to errors in their model simulations, as we show below.
3.4 Agreement concerning the effects of drainage on wetland evaporation
We and Bolivia’s experts agree that the effect of the drainage works in the
immediate area of the Cajones and Orientales springs in Bolivia is to lower the
water table in the area of the springs.
The effect of the drainage works will be to reduce the areas where surface water
would have occurred, and from which the rates of evaporation are relatively high.
However, the water tables are still very close to the surface (15-45 cm from the
Bolivian data), so that water remains readily available for the wetland vegetation
to evaporate. Overall, some reduction in evaporation is expected, making more
water available for discharge in the river. DHI (BCM, Vol. 2, p. 303) estimates
this effect to be equivalent to 2 to 3 l/s of river flow. When we prepared our
contribution to Chile’s Memorial, our best estimate (CM, Vol. 1, p. 161) was that
the annual average would change by 1.3 l/s (0.7% of the flow); however,
recognizing the large uncertainty in these estimates, we quoted an upper bound
estimate of 3.4 l/s, or 2% of the flow (CM, Vol. 5, p. 448). We return to this issue
in section 5 below, in the light of recent work by Chile’s scientists, in which
remote sensing data have been used to estimate evaporation from an undisturbed
wetland (the Quebrada Negra) within the Silala basin in Chile, as well as from the
Cajones and Orientales wetlands in Bolivia (Muñoz and Suárez, 2019). However,
it remains the case that we and Bolivia’s experts are in broad agreement
concerning the impacts of drainage on evaporation from the wetlands and that
these impacts are no more than 2% of the current cross boundary surface water
flow. The water evaporated is a small component of the water balance.
110
14
4 POINTS OF SIGNIFICANT SCIENTIFIC DISAGREEMENT
BETWEEN THE PARTIES’ EXPERTS CONCERNING THE
HYDROLOGY OF THE SILALA RIVER
4.1 Effects of wetland drainage on groundwater discharges to the wetlands
As noted above, one effect of the channelization of the Cajones and Orientales
wetlands is to reduce groundwater water table elevations, and we noted the
agreement between the parties’ experts concerning the effects on evaporation.
A further potential effect of the channelization is to influence the groundwater
flows that feed the wetland springs. DHI states that ‘[t]his has increased the
hydraulic gradients, reduced hydraulic resistance through the springs and
increased their discharge’ (BCM, Vol. 5, p. 83). We agree that these are plausible
effects. However, DHI argues that these effects are so large that natural flows
without drainage and channelization would be 30-40% less than the current
situation (BCM, Vol. 2, p. 266). On this point we strongly disagree. The effects
proposed by DHI on groundwater discharges to the wetlands will be small; in our
opinion the very large estimates made by DHI are implausible. We note that the
DHI results are based entirely on their Near Field modelling. While we have yet
to be provided with details of the model configuration, boundary conditions or
parameters, nevertheless from the available summary information we believe that
there are important flaws in the modelling, and that these have led to these
exaggerated effects, as we show below.
We also note that the 30-40% changes in surface flow referred to by Bolivia do
not include the compensating increases in groundwater flow to Chile. As can be
seen from Table 1 of Annex H to the DHI Report (BCM, Vol. 5, p. 67),
reproduced below, in the baseline case, the combined river and groundwater flows
total 256 l/s, and in the ‘no canal’ and ‘restored wetlands’ cases, the combined
outflows are 209 and 207 l/s respectively – i.e. an 18% and 19% reduction in
111
15
total flow to Chile. Bolivia’s exaggerated claim of a 30-40% change in surface
flows is therefore misleading; they are claiming a change in water flowing from
Bolivia to Chile of less than 20%. However, as discussed in more detail below,
this is an error; whether as surface water, or as groundwater, the water from the
Silala River catchment area flows from Bolivia to Chile.
4.1.1 Water balance considerations
A basic reason for not accepting DHI’s estimates of major changes to
groundwater discharge to the wetland springs comes from simple consideration of
the water balance of the Groundwater Catchment (their Hydrological Catchment).
Provided the groundwater catchment remains the same, the recharge to the
aquifer(s) must either emerge from springs, and flow to the Silala River, or flow
as groundwater down the hydraulic gradient toward the lower end of the
catchment, in the process crossing the international border into Chile.
Recharge to the Silala groundwater catchment, which supplies the Bolivian
springs and the groundwater flow to Chile, is independent of the groundwater
flow regime and, unless the recharge area changes, the sum of these flows should
remain the same. Fundamentally, the water recharging the aquifers which supply
the Bolivian springs and seepages in both Cajones and Orientales wetlands must
be accounted for in either flow that emerges from the springs and wetlands to
form the Silala River or as groundwater flow, in this case into Chile, as agreed by
Bolivia’s experts. Any works on channels or in wetland restoration may affect
local evaporation, and the balance between surface flow and groundwater flow,
but will not affect recharge. Recharge to the aquifer system must therefore be
matched by balancing outflows, which might include evaporation, surface flows
or groundwater flows.
These groundwater flows, as agreed by Bolivia’s experts (DHI, 2018a), flow to
Chile in the south west in a regional aquifer composed of Ignimbrite deposits
112
16
which pass underground from Bolivian territory to Chile. The aquifer occupies the
Silala groundwater catchment and extends into Chile. Should hydraulic resistance
at the Bolivian wetland springs increase to reduce surface flow, as proposed by
DHI (2018b), the groundwater would still flow down gradient into Chile, as
explained in Peach and Wheater (2019).
Groundwater recharge to supply the Cajones and Orientales springs is calculated
by DHI as the difference between precipitation and evaporation (estimated by
DHI to be 24 mm/year (BCM, Vol. 3, p. 478)) over their Hydrological Catchment
area of 234 km2 (BCM, Vol. 2, p. 274). As discussed above, this recharge rate is
unaffected by drainage of the wetlands, nor in DHI’s calculations is there any
mention of the recharge area changing. However, the results from DHI’s near
field model clearly show that the inflow to the model, which arises from this
recharge, changes for the different scenarios considered. In Table 1 below
(reproduced from BCM, Vol. 5, p. 67, Table 1), the ‘inflow’ varies from 3116 to
2655 mm/year (253 to 216 l/s flow equivalent), depending on scenario. These
significant inflow changes make no physical sense, and are unexplained by DHI.
They are the primary reason for DHI’s estimates of the changes in surface water
and groundwater outflows to Chile.
113
17
Table 1. DHI’s results of its modelling of different scenarios
(reproduced from BCM, Vol. 5, p. 67, Table 1).
The question then arises, if in the no-channelization and restored wetlands
scenarios surface water and groundwater outflows to Chile are indeed 18-19%
lower than in the baseline scenario with channels, given the strong topographic
and hydraulic gradients directing groundwater to the springs, and
precipitation/recharge remaining the same, where will the ‘extra’ recharge water
go to if not to Chile?2
There is no obvious pathway for recharge to be diverted around the Near Field
model domain (see Peach and Wheater, 2019), so this change in recharge input
2 It is estimated by Bolivia that impacts of drainage on wetland evaporation account for 2-3 l/s
(BCM, Vol. 5, p. 67, Table 1), so evaporation effects cannot account for a change in outflows that
is estimated by DHI (for the Restored Wetlands in comparison with the Baseline Scenario) to be a
49 l/s net difference (i.e. a loss of 60 l/s in surface flows, less a gain of 11 l/s in groundwater flows
into Chile).
Table 1 Summary of key scenario results
Baseline Scenario No canal scenario Restored wetlands
Water balance
Volume Flow Volume Flow Volume Flow
component
equivalent equivalent equivalent equivalent equivalent equivalent
(mmly) (Vs) (mmly) (1/s) (mmly) (Vs)
Inflow 3116 253 2722 221 2655 216
Storage change 49 4 12 1 64 5
Evapotranspiration 125 10 150 12 164 13
Error 25 2 0 0 -2 0
Outflow (canals) 1846 150 0 0 0 0
Outflow ( overland) 0 0 1159 94 1112 90
Outflow (groundwater) 1310 106 1418 115 1441 117
114
18
must be regarded as unrealistic. Although Bolivia has not provided full details of
the model configuration or boundary conditions, we believe that this is most likely
a direct result of the boundary assumptions in DHI’s modelling, which we discuss
in sections 4.1.2 and 4.1.3 below in more detail.
4.1.2 Increased hydraulic gradients
Turning to the effects of the channelization in reducing water levels in the
wetlands, we recall DHI’s statement (BCM, Vol. 5, p. 83) that the drainage has
increased the hydraulic gradients. It should be noted that one of the most basic
laws of groundwater flow is Darcy’s law (Darcy, 1856), which states that
groundwater flow rate is proportional to the gradient of groundwater potential
energy, or head.3 So a change in gradient can indeed be expected to generate a
change in groundwater flow rate.
However, we recall that DHI (BCM, Vol. 2, p. 280, Figure 6) shows that the
collector drains in the Cajones (Northern) wetland are generally less than 50 cm
deep, and that the measured water tables in both wetlands are between 15 and
45 cm below surface. This is a very small change in water table elevation.
We note that DHI has defined a groundwater recharge area in Bolivia that extends
up to some 20 km from the springs and to topographic elevations of
approximately 5686 m.a.s.l., compared to the spring elevations of 4370 m.a.s.l.,
i.e. a 1316 m topographic difference. We have no information on groundwater
levels in this distant recharge area, but we use DHI’s results (BCM, Vol. 3, p. 488,
Figure 11) (adapted below as Figure 2) to explain our concerns.
3 For 1 dimensional flow in the s direction, Darcy’s law states that (see, e.g., Verruijt, 1970):
q = -KA dh/ds,
where groundwater flow rate is q (m3/d), K is the hydraulic conductivity of the aquifer (m/d), A is
the cross-sectional area of flow (m2), and h is the potential energy, or head (m).
115
19
Figure 2. Simulated groundwater potential (head) in the Ignimbrite aquifer, adapted from
a DHI’s figure (BCM, Vol. 3, p. 488).
It can be seen from their simulated groundwater elevations and their cross section,
shown here as BC, that groundwater flows from a groundwater elevation (head) of
approximately 4520 m.a.s.l. to the Cajones and Orientales springs at
approximately 4370 m.a.s.l. (i.e. a vertical difference of 150 m), over a horizontal
distance of approximately 10500 m. It is therefore difficult to conceive how a
lowering of water table elevation by less than 50 cm (i.e. less than 0.3% of the
groundwater elevation difference of 150 m) can significantly affect groundwater
discharge.
DHI produced these estimates using a very small-scale model (their Silala ‘Near
Field’ model) of the immediate vicinity of the springs (BCM, Vol. 5, p. 16,
lnac:aliri Poli<e
Station
CODELjlntake
Kilomerers
~ator ?rojectiol,WGS84
ILAilNE"' R FIELD
r .
l -1 i-1 y.
BRAZIL
l 'Z"RAGUAY
ARGENTINA )_
_,.
116
20
Figure 2). The area represented is 2.56 km2, approximately 1% of their
Hydrological Catchment area of 234 km2.4 DHI emphasize that their results are
subject to high levels of uncertainty, hence ‘quantitative uncertainty analysis is
not feasible’ and further note that ‘model uncertainty should not be ignored in the
interpretation of results’ (BCM, Vol. 2, p. 303). However, there has been no
attempt made to consider model uncertainty. Further, there are several basic
problems with the model, which lead us to conclude that the model set-up, and in
particular the boundary conditions, explains the erroneous results.
When setting up a groundwater model, a key issue is the selection of the
conditions at the model boundaries (known as the boundary conditions). For
catchment simulation, it would be common to take the whole basin as the
modelled domain, typically with no-flow boundaries, so that recharge is included
in the simulation. However, as DHI use a small Near Field domain, an alternative
approach is needed to represent the boundary conditions. This could be a specified
inflow at the boundary, a specified head, or some combination of the two. As
noted in a standard text book on groundwater modelling by Rushton and Redshaw
(1979, p. 132), ‘when specifying a groundwater flow problem it is common
practice to take a line along which the groundwater potential is constant, and to
enforce this as a boundary condition. This is a valid condition if the groundwater
potential remains at this constant value because the aquifer is in hydraulic
continuity with the sea or a large lake. However a fixed potential implies that
there is an infinite source of water on which the aquifer can draw.’5
DHI have used a ‘fixed head’ boundary condition to represent groundwater flow
into the model from the groundwater recharge area (their hydrological catchment)
(BCM, Vol. 5, p. 18). This means that they have fixed the water table elevation
4 This is a reasonable approach in principle, but only if the flows at the boundaries of the model
can be correctly represented. Conventionally such a small area model would be nested within a
larger scale model to overcome this difficulty.
5 Emphasis added.
117
21
(and hence the groundwater ‘head’ or potential energy) at the boundary of their
model, which is very close to the springs. They then vary the water table in the
wetland, to allow for the effect of drainage, but the upslope boundary water table
remains unchanged, whereas in reality it would also respond to the change in
wetland water table (a water table rise/fall in the wetlands would be accompanied
by a water table rise/fall at the near field boundary). This violates Rushton’s basic
criterion that a fixed potential is a valid boundary condition if the groundwater
potential remains at a constant value. Rushton also notes that a fixed potential
implies that there is an infinite source of water on which the aquifer can draw. In
this case, the fact that while the boundary head is fixed, the inflow to the model is
unconstrained, has allowed the inflow to the model to change, in response to the
changing head gradient due to the change in wetland water level. In reality, of
course, the boundary inflow is equal to the amount estimated from the recharge
calculation. DHI’s choice of a fixed head boundary condition has given rise to the
change in ‘Inflow’ noted above. The incorrect boundary condition thus gives rise
to the incorrect changes in inflows in DHI’s Table 1.
This boundary condition is unrealistic, and this has important consequences
because, as noted above, a) the water table gradient change determines the
groundwater flow,6 b) the gradient change is exaggerated using this assumed fixed
water table so close to the wetland, and c) the fixed head boundary condition
imposes no constraints on the rate of flow across the boundary, so the inflow
changes to accommodate the errors associated with the exaggerated gradient
change.
6 Darcy’s law (Darcy, 1856) states that groundwater discharge is proportional to the gradient of
groundwater potential energy, or head.
118
22
A simple topographic cross-section illustrates the fact that this boundary condition
grossly exaggerates the effect of the drains on the gradient of the groundwater
flow. We use as an example, based on DHI’s simulated heads, a hypothetical flow
path AB, from the springs to the Near Field model boundary (Figure 3a),
a distance of around 360 m, to show that a change of 0.5 m in water table
elevation at the springs changes the average gradient of groundwater head by
0.0794 degrees.
However, if we refer back to the far field flow path of Figure 2, and consider a
fixed head to be specified at the far field boundary, where it is more likely to be
constant, rather than at the near field boundary, the change in average gradient is
0.0027 degrees (Figure 3b).
The average gradient differs by a factor of 29. It is obvious from Darcy’s law that
this will have important consequences in calculating flows, and it can therefore
readily be appreciated, from even this simple geometric comparison, that the
assumption of the fixed head at the near field boundary has important
consequences for DHI’s calculation of the effect of the drains on the spring flows,
which we discuss below.
Figure 3a. A typical groundwater head gradient from the near field model
boundary to the wetland.
Figure 3a. A typical groundwater head gradient from the near field model
boundary to the wetland.
Figure 3b. A typical groundwater head gradient from the far field model
boundary to the wetland.
23
Figure 3a. A typical groundwater head gradient from the near field model
boundary to the wetland.
Figure 3b. A typical groundwater head gradient from the far field model
boundary to the wetland.
119
300 600
Meters
Mere,ator Projection, WGS 84
/
FCABI
)
:[ 4380
C
0 ·;::;
">'
QJ
LU
4370
43600 so
900
CHILE
BOLIVIA
Area
Enlarged
/
Change of groundwater gradient using Bolivia's near field boundary
100 150 200
Distance (m)
250
0
300 350 400
' 1
L
lnacaliri Police
Stati~
0 4
Kilometers
Mercator Projection, WGS 84
Hilo S/N-Ll(XII
FCAB
Intake <>
Rio Si\O~
9.,
..i,. ebractq
CODELCO Intake egro HitDS/N
o~ , H ~ l "<>- Q ,,
i Q ""' I ~
6
,s
7 ' o 1 V
Change of groundwater gradient using Bolivia's far field boundary
4600 --------------------------------------------
:[
C
0 ·;::;
"if;'
LU
4500
43000
1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 11000
Distance (m)
24
121
25
A text book calculation, based on an analytical solution to the equations of
groundwater flow, can be used as an example to indicate the potential effect of the
erroneous boundary condition on the groundwater discharge in a little more detail
(see Appendix 1 of this report for details and, for example, Verruijt (1970), p. 53).
This is a two-dimensional calculation, based on uniform properties, and is not
therefore attempting to represent the detail of the DHI simulation, which is not
known to us, but merely to demonstrate the magnitude of the effect of their
boundary condition assumptions. We consider first the effect of a change in water
table elevation in the wetland, due to channelization. Later we consider a change
in resistance to flow.
Figure 4 shows an idealized two-dimensional hillslope segment7 of length 360 m
based for illustration on the Near Field model section of Figure 2, draining from a
constant head boundary (at 4385 m.a.s.l.) to discharge to a wetland (at
4370 m.a.s.l.). The difference in head is 15 m. For the purposes of demonstration,
we assume that the aquifer is uniform, with a typical hydraulic conductivity
(4.3 m/day, or 5x10-5 m/s, see BCM, Vol. 5, p. 21, Table 3, Upper Silala
Ignimbrite), recharge rate (24 mm/year, BCM, Vol. 3, p. 478) and an aquifer
depth of 400 m (BCM, Vol. 5, p. 17), as used by DHI. If the water table is
increased by 0.5 m to represent the effect of infilling the channels with aquifer
material, the groundwater discharge decreases by 3.3%. Alternatively, if we take a
fixed head boundary condition at the far field boundary, as in Figure 2,
section AC, 10500 m away (Figure 5), and increase the water table at the springs
by the same 0.5 m, the effect is 0.28% decrease in discharge. Both effects are
small, but DHI’s incorrect choice of boundary condition exaggerates the effect of
hydraulic gradients on water table change due to channelization by a factor of 12
(note that this is less than the factor of 29 quoted above due to the additional detail
included in these calculations; specifically the recharge applied along the section,
7 Note that scales are distorted to allow the problem to be visualized.
122
26
which results in a groundwater gradient that varies along its length, unlike in the
simpler example above).
Figure 4. Effect of fixed groundwater head at near-field boundary on spring discharge.
Figure 5. Effect of fixed groundwater head at a more realistic 10.5 km boundary on
spring discharge.
h0 = 415m
h0 = 550 m
Recharge = 24 mm/year
m
K = 4.3 -
day
Unconfined Aquifer
mpermeab e Layer
L = 360 m
hL = 400 m
Recharge = 24 mm/year
m
K= 4.3-d
ay
Unconfined Aquifer
mpermeab e layer
L = 10,500 m
hL = 400 m
(+ O.S m)
.....
Channel
(+O.Sm)
.....
Channel
Flow with channels:
73.0~
day•m
Flow w/o channels:
70.6~
day•m
6q=2.4~
day·m
3.3% decrease
Flow w ith channels:
29.52~
day -m
Flow w/ o channels:
29.44~
day·m
6q = 0.08 ~
day-m
0.28% decrease
123
27
4.1.3 Reduced hydraulic resistance
A second effect of the drainage and channelization works proposed by DHI
(BCM, Vol. 5, p. 83) is that of reduced hydraulic resistance. DHI argues that
construction of the drainage works has removed a layer of soil8 (BCM, Vol. 2,
p. 374), and that this has increased the groundwater discharge. They also argue
that in a restored wetland, up to 60 cm thickness of peat will develop,
accumulating at a rate of 0.1-1 cm/year, and that a thicker peat layer implies a
higher resistance to groundwater emerging in the wetlands (BCM, Vol. 5, p. 70).
There are several reasons why this is a misleading simplification of the
hydrogeological situation, as discussed in section 5 below, see also Muñoz and
Suárez (2019). However, once again an elementary calculation (Appendix 1,
part 2) shows that the choice of the Near Field model boundary condition grossly
exaggerates any such effect. Figure 6 shows a near field fixed head boundary,
360 m from the spring emergence, with the same elevation difference (15 m) as
above. We introduce a ‘buffer zone’ at the base of the slope, representing
conceptually the effect of a 1 m layer of peat on the groundwater flow path
adjacent to the channel. We take peat permeability 2 orders of magnitude lower
than the ignimbrite aquifer (0.043 m/day = 5 x 10-7 m/s, consistent with the lower
limit of DHI’s assumptions (BCM, Vol. 5, p. 26, Table 4)). The effect of the
buffer zone is to reduce the groundwater inflow to the wetland by 22%. However,
if a far field (10500 m) fixed head boundary is used (Figure 7), the effect of the
same buffer configuration is a flow decrease of just 0.9%. The buffer zone has a
disproportionate effect (by a factor of 23) on the flow field in the Near Field
model, due to the choice of near field fixed head boundary.
If we now combine the effect of changing hydraulic gradient and reduced
hydraulic resistance and superimpose the effect of a 0.5 m water table rise
8 DHI argues that ‘By excavating the soil […] the hydraulic resistance to the groundwater
discharge […] has been reduced.’
124
28
together with the peat layer (also shown in Figures 6 and 7), the flow decreases by
24% for the near field configuration, and 1.2% for the far field, i.e. the effect is
magnified by a factor of 20.
While these calculations are highly simplified, they clearly demonstrate that the
inappropriate choice of near field boundary condition has grossly exaggerated the
effects of the drainage channels, by around 20 times, on both the reduced
groundwater heads and any hypothetical reduced hydraulic resistance. The Far
Field calculation of a 1% effect of these changes is indicative of the expected
order of response, and when combined with possible changes in evaporation (and
also considering the water balance issues discussed in Section 3.1.1), we remain
of the view that the impact of the channelization on river flows is of the order of a
few percent change in river flows.
Figure 6. Effect of low conductivity zone on spring discharge – near field fixed head.
h0 = 415 m
Recharge = 24 mm/year
m
K1 = 4.3d
lgnimbrite Aquifer
Unconfined Aquifer
lmpermea le Layer
L = 360 m
m
K2 = 0.043d
Peat Layer
hi = 400 m
(+O.Sm)
Channel
Floww/o peat layer
(d2=0) and no water
table rise:
73.0~
day•m
Flow with peat layer
(d2: l m) and no water
table rise:
57.3~
..... d ay-m
Flow with peat layer
(d2=1 m) and 0.5 water
table rise:
55.4~
d ay-m
~q = 17.6 da:~m
24% decrease
125
29
Figure 7. Effect of low conductivity zone on spring discharge – far field fixed head.
4.1.4 Other modelling issues
As noted above, Bolivia’s assertions concerning the impacts of channelization
depend entirely on the use of a Near Field coupled surface-groundwater model to
simulate the effects. We have shown above that a basic error in the choice of
model boundary condition has led to the exaggeration of these effects, and that
resulting change in recharge inputs to the model is incorrect. While we have
focused above on incorrect specification of the near field model inflow boundary,
a similar problem arises with other model boundaries. For example, the near field
model outflow boundary is specified as a fixed hydraulic gradient, which makes
no allowance for the fact that a changing groundwater flow gradient would be
expected in response to the hypothesized flow changes. Also the groundwater
table elevation contours shown in Figure 35 of Annex G to the DHI Report are
inconsistent with the model’s assumed lateral boundary conditions (BCM, Vol. 5,
p. 49). We believe these are no-flow boundaries, but if that is the case, the
contours of groundwater head must be orthogonal to the model boundaries, and
clearly that is not the case in this Figure.
h0 = 550 m
Recharge = 24 mm/year
m
K1 = 4.3d
lgnimbrite Aquifer
Unconfined Aquifer
lmpermea le l ayer
L = 10,500m
(+O.Sm)
hi= 400 m
Channel
Flow w/o peat layer
(d2=0) and no water
table rise:
29.52 ~
day•m
Flow with peat layer
{d2=1 m) and no w ater
table rise:
_, 29.25~
____,......- aay•m
Flow with peat layer
(d2=1 m) and 0 .5 water
table rise:
29.17~
day•m
~q = 0 .35 d;~m
1.2% decrease
126
30
Both we and Bolivia’s experts understand that the hydrogeology of the Silala
system is complex, and difficult to understand from the limited data available.
Nevertheless, the detail of the local geology is particularly important in
understanding the Near Field groundwater flows, and any response to
channelization. The DHI groundwater representation is flawed in many important
respects, and not representative of the true situation, as discussed in detail in
Peach and Wheater (2019). One obvious indicator of this is that the groundwater
modelling fails to recognize the fact that both the major ion and carbon isotope
chemistry of the Cajones and Orientales spring waters are very different, and
therefore the sources of the water are different. The DHI Hydrogeological
Conceptual Model recognizes this (BCM, Vol. 2, p. 294), but the numerical
modelling fails to take into account this basic feature of the data. This and other
aspects of the geology are discussed in more detail by Peach and Wheater (2019).
It is also relevant to note at this point that there are other areas of concern with the
model, although our analysis is hampered by the fact that (as yet) we have no
detailed information on the configuration, parameters, forcing data or outputs of
the various models used. Nevertheless several issues are immediately evident:
1. The near field model is stated to have been run as ‘steady-state’ (BCM,
Vol. 5, p. 13, ‘a stationary model approach has been adopted’). Since there
is no variation with time, this means that the model inputs must equal the
model outputs, with, by definition, no change in storage. This is not the
case in the DHI results. For example in DHI’s summary table of results
(BCM, Vol. 5, p. 67, Table 1) annual storage changes (accumulating each
year) are quantified for each of the 3 scenarios (baseline, no canal, restored
wetlands). This is a basic model error.
2. The near field model results as presented in the same table do not add up.
For example under the Baseline Scenario, the inflow to the model (from
recharge) is 253 l/s flow equivalent, and the losses from the model total
127
31
266 l/s (loss to evaporation 10 l/s, surface water outflow 150 l/s and
groundwater outflow 106 l/s). Even allowing for an ‘error’ term (2 l/s) and
a change in storage (4 l/s, but note this term is not permissible in a steady
state model), the numbers do not make sense. Also under the Restored
wetlands Scenario, the inflow (216 l/s) and the total outflow/losses
(220 l/s) are not the same.
4.2 Other areas of disagreement
There are many additional points of detail for which DHI’s analysis differs
somewhat from ours, including for example precipitation over the Silala
topographic basin and its larger groundwater catchment. However, in our view,
these other differences are relatively minor in the overall context of this dispute,
and in many respects to be expected, given the challenges of quantifying
hydrological response with very limited data. However, the key point is that the
DHI Near Field model is based on inaccurate geology (as discussed in detail by
Peach and Wheater, 2019), has inappropriate boundary conditions, resulting in
inconsistent water balances, and therefore does not explain correctly the effects of
the channelization in Bolivia on the surface and groundwater flows to Chile.
5 NATURAL VARIABILITY AND FUNCTIONING OF THE
BOLIVIAN WETLANDS AND TOPOGRAPHIC CONSTRAINTS
The BCM refers in many places to wetland degradation without providing any
supporting data to show that significant degradation has occurred over time. To
the contrary, Bolivia cites results from Castel (2017) that confirm Chile’s
observations (CM, Vol. 4, p. 37, Figure 16) of the strong role of natural annual
and seasonal variability in determining the area of active wetland vegetation, but
show no long term trend over the recent decades for which data are available.
128
32
Despite that, DHI refers to the changes made by channelization in the Bolivian
wetlands and asserts, for example, that ‘[t]his has reduced the size of wetlands’
(BCM, Vol. 2, p. 374). The Ramsar report (BCM, Annex 18) states that ‘[a]t
present, there are only vestiges of the original wetlands that used to cover an area
of about 141,200 m2, or 14.1 hectares. The current surface area of the wetlands
covers only about 6,000 m2, or 0.6 ha. […]’ (BCM, Vol. 5, p. 163). This is clearly
a wildly inaccurate statement, as can be seen from a) Bolivia’s own data of the
wetland areas (Castel, 2107), and b) Chile’s remote sensing data, presented in the
Memorial (CM, Vol. 4, p. 37, Figure 16), in Chile’s Reply (Muñoz and Suárez,
2019), and in summary data in this report, see Table 2 below. It is not our
intention to state that the channelization is without any effect on the Bolivian
wetlands, however it is important that any degradation be appropriately quantified
so that its effects on the Silala River flow may be better understood.
To aid understanding of the functioning and dynamic behaviour of the Silala
wetlands, Chile has recently established a detailed study of the Quebrada Negra
wetland, located within the Silala River topographic basin in Chile (Muñoz and
Suárez, 2019), as shown in Figure 8. This wetland is of comparable areal extent to
the Cajones and Orientales wetlands in Bolivia (approximately 3 hectares) but is
undisturbed by any human activity.
129
33
Figure 8. Location of the Quebrada Negra wetland within the Silala River topographic
catchment in Chile (Muñoz and Suárez, 2019).
Selected photographs of the wetland are reproduced here, in Figures 9-11, after
Muñoz and Suárez (2019). It can be seen that the wetland fills the valley bottom,
which is characterized by natural channels that flow in response to spring
emergence, interconnect to form a braided network, then lose their flow through
re-infiltration. It can also be seen that vegetation extends up the base of adjacent
slopes, focused on small tributary ravines, indicating spring emergence at the
hillslope boundaries of the lowland wetland.
,~o <o \iri
\,I>
tt
r5
COInOtEa~LkCeO •
\ ~
lnaL,i .
Police Station
'(r Mel<.l!CrPr~ection.WGS84
i910 Silo/a /
Quebrada
Negra
C
Hito
5/HFCAB
I
Quebrada Negra
study area
□>-
Military
Post
•I
Hlto
S/ N-LXIY
HitoS/N
130
34
Figure 9. Photograph of the Quebrada Negra wetland, taken from the northern slope
(Muñoz and Suárez, 2019).
Figure 10. Photograph taken at the Quebrada Negra wetland, looking upstream
(Muñoz and Suárez, 2019).
131
35
Figure 11. Photograph of the Quebrada Negra wetland, taken from the southern slope
(Muñoz and Suárez, 2019).
The visual images from Bolivia in the Counter-Memorial show extensive areas of
active vegetation in the Cajones and Orientales wetlands, despite the presence of
the channelization (e.g. BCM, Vol. 2, p. 273, Figure 4; p. 333, frontispiece;
p. 370, Figure 5; p. 372, Figure 8). See, for example, the frontispiece of Annex B
to the DHI Report, reproduced as Figure 12, below:
Figure 12. Photograph of Bolivian wetland (BCM, Vol. 2, p. 333).
132
36
Chile’s Memorial used Landsat remote sensing imagery (from 1987 to 2016) to
show that the active area of the Cajones and Orientales wetlands varied strongly,
both seasonally and from year to year (CM, Vol. 4, p. 37, Figure 16), but with no
overall trend. This point is also made by Bolivia. Castel (2017) uses Landsat
imagery from 1975 to 2000 to come to the same conclusions, as noted above –
high seasonal and inter-annual variability is seen in the wetland vegetated area, as
mapped using the Normalized Difference Vegetation Index NDVI (which shows
vegetation activity), but no evident trend.
In Chile’s recent work, Muñoz and Suárez (2019) use higher resolution satellite
imagery (Sentinel-2, 10 m resolution) for the period July-November 2018
(Figure 13), which allows the extent of vegetation extent to be mapped onto the
topography.
133
37
Figure 13. Quebrada Negra, Cajones and Orientales wetlands average NDVI distribution
from July to November 2018 (Muñoz and Suárez, 2019).
Normalized Difference Vegetation Index NOVI
D < = 0.20 • 0.20 - 0.25 D 0.25 - 0.30 D 0.30 - o.35 , o.35
134
38
The results are reproduced in Figures 14-16 below for the Bolivian Cajones and
Orientales wetlands, and the Chilean Quebrada Negra. It can be seen that the
active vegetation fully occupies the lowland areas of the respective valleys, and
that seasonally, vegetation expands up the base of the adjacent slopes.
Figure 14. Cross section of vegetation cover (NDVI>0.2) and topography of the
Quebrada Negra wetland. Average (Green Line) and Maximum (Red Line)
cross section of vegetation cover have the same extension. For this reason,
only average green cover (green line) is visible (Muñoz and Suárez, 2019).
4307
4302
4297
14292
C
0 ·;;
"~' 4287
w
4282
4277
4272
Cross Section Quebrada Negra
NDVIJul/18-Nov/18 =0.2
0 25 so 75
Cover vegetation estimation using
Sentinel-2 0
100 125
200
Average section covered by
vegetation (NDVl=0.2, G
Maximum section covered by
vegetation (NDVl=0.2, Red line)
between Jul/18-Nov/18
- AW3DTM
- Max
- Average
150 175 200
Horizontal distance [m]
135
39
Figure 15. Cross section of vegetation cover (NDVI>0.2) and topography of
the Cajones wetland (Muñoz and Suárez, 2019).
Figure 16. Cross section of vegetation cover (NDVI>0.2) and topography of the
Orientales wetland (Muñoz and Suárez, 2019).
4383
4378
~ 4373
E"'
5 4368
.:;
">'
Q)
u:i 4363
4358
4353
Cross Section Cajones
NDVIJul/18-Nov/18 =0.2
0 25 50 75
Cross Section Orientales
NDVIJul/18-Nov/18 =0.2
4422
4420
4418
.,;
1 4416
C:
0
~ 4414
>
~
LU
4412
4410
4408
0 25 so
Cover vegetation
Average section covered by vegetation
(NDVl=0.2, Green line)
Maximum section covered by vegetation
(NDVl=0.2, Red line)
- AW3DTM
- Max
100 125 150 175 200 225 250 275 300
Horizontal distance [m]
Cover vegetation estimation using
Sentinel-2
75 100 125
Horizontal distance [m]
200
Average sect ion covered by
vegetation (NDVl=0.2, Green line)
Maximum section covered by
vegetation (NDVl=0.2, Red line)
between Jul/18-Nov/18
- AW3DTM
- Max
,--------' - Average
150 175 200
136
40
The conclusion is that the channelization activities in Bolivia’s wetlands, which
are focused entirely on the flat topography of the valley floors, have not
significantly affected the area of active wetland in the valley floors.
We turn now to the effects of channelization on wetland evaporation. While we
expect some impacts of the drainage works on wetland vegetation, as areas of
surface water ponding will have been reduced, the fact that the water tables
remain so close to the surface means that wetland vegetation has a plentiful water
supply and can transpire freely.
The key indicator is obtained from the high resolution Sentinel-2 satellite NDVI
data, for the period July-November 2018, summarized in Table 2 below (after
Muñoz and Suárez, 2019). We recall that this period is the southern hemisphere
Winter, emerging into Spring. Note that the combined area of the Cajones and
Orientales wetlands is always much larger than the 0.6 hectares quoted as the
‘current’ area in the Ramsar report (BCM, Vol. 5, p. 163).
Area covered by vegetation (ha)
July August September October November
Cajones wetland 0.81 1.12 1.31 2.2 2.41
Orientales wetland 2.23 2.7 2.86 6.09 7.5
Combined Cajones and
Orientales wetlands 3.04 3.82 4.17 8.29 9.91
Quebrada Negra wetland 2.13 2.31 2.58 4.12 3.43
Table 2. Area covered by vegetation in the Quebrada Negra, Cajones and Orientales
wetlands, from July to November, 2018 (Muñoz and Suárez, 2019).
A first order estimate of the annual actual evaporation rates can be derived from
summer NDVI data, following the methodology of Groeneveld (2007), as
explained in Muñoz and Suárez (2019). The results are shown in Table 3, below.
137
41
Annual ETa,NDVI,
(mm/year)
Mean S.D.
Cajones wetland 705 17
Orientales wetland 702 23
Quebrada Negra wetland 631 21
Table 3. Annual ETa,NDVI, Mean and Standard Deviation (S.D.) in mm/year estimated for
the Quebrada Negra, Cajones and Orientales wetlands (after Muñoz and Suárez, 2019).
The estimated annual evaporation rates from the Cajones and Orientales wetlands
are very similar, and 10% higher than the evaporation rate from the Quebrada
Negra wetland. Considering the respective wetland areas, the total annual
evaporation is equivalent to a flow rate in the Silala River of 0.6 l/s for the
Cajones wetland, 2.3 l/s for the Orientales and 0.7 l/s for the Quebrada Negra.
As noted above, we and Bolivia’s experts are in broad agreement concerning the
impacts of drainage on evaporation from the wetlands and that these impacts are
no more than 2% of the current cross boundary flow. The water evaporated is a
small component of the water balance. However, the conclusions from the recent
work of Muñoz and Suárez (2019) are that any effects of canalization on wetland
evaporation in the Orientales and Cajones wetlands are non-detectable from the
satellite data, and in fact both Bolivian wetlands appear to evaporate at a greater
rate than the ‘undisturbed’ Quebrada Negra wetland, although this is within the
expected margin of error for this method.
Having established that the channelization of the wetlands in Bolivia does not
seem to have affected the areal extent of the wetlands, or to have had a significant
effect on evaporation rates, we turn again to the field studies reported by Muñoz
and Suárez (2019) to provide further insights into wetland function. A very
detailed groundwater monitoring programme (82 monitoring points, with
groundwater head measured at two different depths at each location, see
Figure 17) has been put in place.
138
42
Figure 17. Layout of the monitoring wells in the Quebrada Negra wetland. Location with
sensors for continuous groundwater level monitoring are depicted in blue
(Muñoz and Suárez, 2019).
Results to date have shown that the areas of groundwater inflow to the wetlands
are quite heterogeneous. In fact, over much of the valley floor, the hydraulic
gradients show downwards flow (Figure 18) – and clearly in such areas, a
drainage channel would not affect groundwater emergence. Areas of upwelling
arise at the upper boundary of the wetland, along the base of the hillslopes that
surround the wetland, and only in limited locations within the main wetland itself.
This perhaps explains why the apparent effects of the drainage channels on the
Bolivian wetlands, in terms of spatial extent, wetland function, and evaporation,
have been more limited than might have been expected by Bolivia.
Meters
PSADS!i/ ITTMzont 19S
139
43
Figure 18. Contour lines of groundwater levels (m.a.s.l.), at shallow piezometers,
measured during September 2018 at the main grassland of the Quebrada Negra wetland.
Red circles represent points where positive gradient (downwelling) was observed, blue
circles represent the points where negative gradient (upwelling) was observed and yellow
circles represent points where zero-gradient was observed. Surface channels observed in
the wetland are identified as light blue lines. Apparent surface water sources are marked
with dashed yellow lines, and C1-C4 are vegetation cover types as defined by Muñoz and
Suárez (2019).
6 CONCLUSIONS
(i) What are the major points of scientific agreement between the Experts of
Bolivia and those of Chile concerning the hydrology of the Silala River?
We and Bolivia’s experts agree that:
1. The Silala River flows naturally from Bolivia to Chile. The river rises in
two sets of springs in Bolivia, which maintain the Cajones and Orientales
wetlands.
2. The river is primarily fed by groundwater and interacts with groundwater
along its course to the border and beyond.
Vertical gradient direction
• Upwelling
o Zero gradient
• Downwelllng
-- Surface water channels
10 20
Meters
140
44
3. In addition, there are substantial groundwater flows from Bolivia to Chile,
likely of an equivalent magnitude to the surface water flows.
4. Construction of drainage channels and river channelization in the 1920s will
have had some effect on the flow. An increase in flow due to these works is
expected.
5. The impact of drainage on evaporation from the wetlands is small.
(ii) What are the major points of scientific disagreement between the Experts
of Bolivia and those of Chile concerning the hydrology of the Silala River?
We and Bolivia’s experts disagree about the magnitude of the impact of the
drainage works. In our opinion, Bolivian estimates of a 30-40% effect on flows
are implausible. These estimates have been produced by a Near Field model of
surface water-groundwater interactions. We have shown that the model is based
on incorrect geology, that simple calculations show that incorrect assumptions of
the model’s boundary conditions lead to an overestimate of the impacts, by a
factor of approximately 20, and that the change in inputs to the model is
unrealistic.
(iii) What new evidence has been produced, since Chile submitted its Memorial
in July 2017, concerning the effect of the channelization of the flow on Bolivian
territory on the watercourse of the Silala River that flows from Bolivia into Chile?
New studies based on detailed monitoring of an undisturbed Chilean wetland
within the Silala basin, coupled with high resolution remote sensing data, show
that Bolivian and Chilean wetlands continue to fully occupy the valley floor, and
seasonally extend up the base of adjacent hillslopes. The condition of the wetland
vegetation, as indicated by remote sensing, is similar in all three wetlands, and
associated estimates of actual evaporation suggest that the higher evaporation
141
45
rates are observed from the Cajones and Orientales wetlands, some 10% greater
than that of the undisturbed Quebrada Negra wetland. At least from the satellite
data, it appears that there has been no significant change in evaporation associated
with the channelization of the Bolivian wetlands, and hence no effect of
evaporation change on river flows.
In summary, we remain confident that the effects of the drainage works on
evaporation are quite limited, as stated in Chile’s Memorial, at most equivalent to
a flow of 2-3 l/s on average, i.e. some 2% of the natural flow, but in the light of
our recent results, probably less. Other effects will be similarly small. Bolivia’s
estimates of 30-40% changes in river flow are due to errors in DHI’s modelling
and are implausible. We also reiterate that there is no doubt that the Silala River is
an international watercourse, and we are pleased to note the agreement of
Bolivia’s experts on this point.
142
46
7 REFERENCES
Alcayaga, H., 2017. Characterization of the Drainage Patterns and River
Network of the Silala River and Preliminary Assessment of Vegetation Dynamic
using Remote Sensing. (Chile’s Memorial, Vol. 4, Annex I).
Castel, Ana Paola. Análisis Multitemporal mediante imágenes de satélite de los
Bodefales de los Manantiales del Silala, Potosí – Bolivia. DIREMAR. La Paz.
(Chile’s Reply, Vol. 2, Annex 98).
Danish Hydraulic Institute (DHI), 2018a. Study of the Flows in the Silala
Wetlands and Springs System. (Bolivia’s Counter-Memorial, Vol. 2, Annex 17).
Danish Hydraulic Institute (DHI), 2018b. Study of the Flows in the Silala
Wetlands and Springs System. (Bolivia’s Counter-Memorial, Vol. 5,
Annex 17-I).
Darcy, H., 1856. Les fontaines publiques de la ville de Dijon. Dalmont.
Dupuit, J., 1863. Etudes théorique et pratiques sur le movement des eaux dans les
canaux découverts at á travers les terrains perméables, 2nd ed., Dunod.
Forchheimer, P., 1886. Über die Ergiebigkeit von Brunnen-Anlagen und
Sickerschlitzen. Zeitschr. Archit. Ing. Ver. Hannover, 32, 539-563.
Muñoz, J.F. and Suárez, F., 2019. Quebrada Negra Wetland Study. (Chile’s
Reply, Vol. 3, Annex XIII).
Peach, D.W. and Wheater, H.S., 2017. The Evolution of the Silala River,
Catchment and Ravine. (Chile’s Memorial, Vol. 1).
Peach, D.W. and Wheater, H.S., 2019. Concerning the Geology, Hydrogeology
and Hydrochemistry of the Silala River Basin. (Chile’s Reply, Vol. 1).
Rushton, K.R. and Redshaw, S.C., 1979. Seepage and Groundwater Flow. Wiley.
Verruijt, A., 1970. Theory of Groundwater Flow. Macmillan.
Wheater, H.S. and Peach, D.W., 2017. The Silala River Today – Functioning of
the Fluvial System. (Chile’s Memorial, Vol. 1).
143
47
APPENDIX 1
SIMPLIFIED CALCULATION OF HILLSLOPE GROUNDWATER
FLOWS
A simplified analytical calculation of groundwater flow in an idealized hillslope is
used to demonstrate that DHI’s use of a fixed head boundary condition for the
Near Field modelling of the Bolivian wetlands leads to exaggerated impacts on
groundwater flows to the wetland, in response firstly to changes in wetland water
table elevation and secondly to the presence of a peat layer of reduced hydraulic
conductivity.
1. Groundwater response to change in wetland water table elevation
We adopt, for demonstration purposes, the groundwater head field proposed by
DHI (BCM, Vol. 3, p. 488, Figure 11) (see Figure A1 below).
144
48
Figure A1. Simulated groundwater potential (head) in the Ignimbrite aquifer,
adapted from DHI’s Annex E, Figure 11 (BCM, Vol. 3, p. 488).
We consider an idealized hillslope, which represents groundwater flow in an
aquifer discharging to a wetland, and consider flow paths of two different lengths,
informed by cross-sections AB and AC on Figure A1. A fixed head boundary
condition at the far field boundary C is compared with DHI’s assumption of a
fixed head at the Near Field boundary B to demonstrate the effect of the boundary
condition assumption on the calculation. The water table elevation in the wetland
is taken as the downstream boundary condition.
We consider a text book calculation (see Verruijt, 1970, p. 53), as shown in Figure
A2 below, of one-dimensional steady groundwater flow, for a cross–section of
unit thickness, under the assumption that flow is approximately horizontal
lnacallrt Police
Kilometers
MmatorP!ojKtkln,WYi 84
Station _ ~~
\., 810 5\\(\ Hilo 16-LXX
\ ~"e¢rqq;
CODELCO Intake <'!ho "
Area
~ 1
Enlarged
~ CHILE
\J ;;:
¥ f }
PARAGUAY ', ARGENTINA 2
...,.
145
49
(the well-known Dupuit-Forchheimer approximation (Dupuit, 1863; Forchheimer,
1886)), which means that the groundwater head can be taken as equal to the water
table height.
Figure A2. Schematic of groundwater hillslope flow, from fixed head upslope boundary to
fixed head wetland water table elevation.
Darcy’s Law relates the groundwater flow velocity through a cross-sectional area
of aquifer to the gradient of groundwater potential, or head.
Hence from Darcy’s law,
𝑣𝑣𝑥𝑥 = −𝐾𝐾 𝑑𝑑ℎ
𝑑𝑑𝑑𝑑
where vx is the groundwater flow velocity (m/day), h is the ‘head’ or water table
elevation (m), K the hydraulic conductivity (m/day) and x the horizontal length
dimension (m).
If we introduce a recharge rate (i.e. precipitation-evaporation) N (m/day), then
from conservation of mass,
h
X
Recharge N
Hydraulic
conductivity K
Unconfined Aquifer
mpermeab e Layer
L
Horizontal distance
Spring Emergence
..... FlowqL
Channel
146
50
𝑁𝑁 = 𝑣𝑣𝑥𝑥
𝑑𝑑ℎ
𝑑𝑑𝑑𝑑+ ℎ 𝑑𝑑𝑣𝑣𝑥𝑥
𝑑𝑑𝑑𝑑
And it can simply be shown that
􁉀𝐾𝐾
2􁉁 𝑑𝑑2(ℎ2)
𝑑𝑑𝑥𝑥2 + 𝑁𝑁 = 0
For a hillslope of length L, with constant head boundaries h0 and hL, if qL is the
discharge per unit width (m3/day/m), known as the specific discharge,
𝑞𝑞𝐿𝐿 = 𝐾𝐾􀵫ℎ0
2−ℎ𝐿𝐿
2􀵯
2𝐿𝐿 + 𝑁𝑁𝑁𝑁
2
For the Near Field calculation, based on the DHI simulation in Figure A1 above
and the flow path from B to A, the difference in heads is taken as the difference
between elevation 4370 m.a.s.l. and 4385 m.a.s.l., i.e. 15 m, with a path length of
360 m. We take DHI’s recharge rate of 24 mm/year (BCM, Vol. 3, p. 478),
hydraulic conductivity K = 4.3 m/day (BCM, Vol. 5, p. 21, Table 3) and a 400 m
deep aquifer below the wetland as proposed by DHI (BCM, Vol. 5, p. 17), and
assume a 0.5 m rise in wetland water table if the channels are removed
(Figure A3).
Figure A3. Schematic for Near Field flow path discharge calculation.
h0 = 415m
Recharge = 24 mm/year
m
K=4.3-
day
Unconfined Aquifer
mpermeab e Layer
L = 360 m
hL = 400 m
(+ O.S m)
Cha nnel
.....
Flow with channels:
73.0~
day •m
Flow w/o channels:
70.6~
day •m
~q=2.4~
day•m
3.3% decrease
147
51
Similarly, for the Far Field, the difference in heads is taken, again from DHI
(2018) and based on flow path CA, as the difference between elevation
4520 m.a.s.l. and 4370 m.a.s.l., i.e. 150 m, with a path length of 10500 m
(Figure A4).
Figure A4. Schematic for Far Field flow path discharge calculation.
The results are summarized in Table A1 below, and show that the incorrect
assumption of a fixed head at the near field boundary increases the effect of a
water table rise by a factor of 12.
h0 = 550 m
Recharge = 24 mm/year
m
K = 4.3 -d ay
Unconfined Aquifer
mpermeab e Layer
L = 10,500 m
hi = 400 m
(+ O.Sm)
Channel
.....
Flow with channels:
29.52--2!!:_
day ·m
Flow w/o channels:
29.44~
aay-m
6q = 0.08~
day -m
0 .28% decrease
148
52
Groundwater discharge
m3/m/day
NEAR FIELD
With channel in place 73.0
With channel removed
0.5 m water table rise
70.6
% decrease in flow 3.3%
FAR FIELD
With channel in place 29.5
With channel removed
0.5 m water table rise
29.4
% decrease in flow 0.28%
RATIO of % CHANGE 11.8
Table A1. Summary of results – channel removal/water table rise.
2. Groundwater response to the presence of a peat layer of reduced
hydraulic conductivity.
DHI assumes that over a long period of time, in the absence of disturbance, an
increased depth of peat will develop in Bolivia’s wetlands (BCM, Vol. 5, p. 70).
As the peat is expected to have relatively low hydraulic conductivity, DHI argue
that this will reduce the groundwater discharge to the wetland. Here, we
demonstrate that if such an effect were to occur, its estimated magnitude would be
grossly exaggerated by the assumption of the fixed head boundary condition in
DHI’s Near Field model. We adopt the same approach as in example 1 above, i.e.
we use a simplified representation of the groundwater flow, which can be solved
analytically, to indicate the nature and potential magnitude of this erroneous
assumption.
For this analysis, we assume that groundwater will flow downslope through a
groundwater aquifer to the wetland area, and emerge though a peat layer that
underlies the wetland, as shown schematically in Figure A5, below. As noted by
DHI (BCM, Vol. 2, p. 279), the depths of peat range typically from 0.2 to 1.0 m in
149
53
the wetlands, and the drainage channels are said to cut through most of the
wetland soils. We therefore assume two cases; one with no peat cover, the other –
a notional undisturbed condition - has 1 m depth of peat.
Figure A5. Schematic representation of hillslope and valley bottom flow paths, showing
near field and far field domains.
We represent the effect of a layer of reduced permeability on the groundwater
flow path by introducing a section of reduced permeability at the base of the
hillslope, as shown in Figure A6, where d2 = 1 m.
Water Tab-le Far Field
Near Field (not in scale)
Bolivian Wetlands
lgnimbrite Aquifer
150
54
Figure A6. Schematic for flow path discharge calculation.
For uniform steady state groundwater flow, with recharge N, the relationship for
specific groundwater discharge 𝑞𝑞𝐴𝐴 in the A domain (left hand zone) is given by:
𝑞𝑞𝐴𝐴(𝑥𝑥1) = 𝑁𝑁𝑥𝑥1 −
𝑁𝑁𝑑𝑑1
2
+
(ℎ0
2 − ℎ1
2)
2𝑑𝑑1
𝐾𝐾1
In addition, the specific groundwater discharge 𝑞𝑞𝐵𝐵 in the B domain (right hand
zone) is given by:
𝑞𝑞𝐵𝐵(𝑥𝑥1) = 𝑁𝑁(𝑥𝑥1 − 𝑑𝑑1) −
𝑁𝑁𝑑𝑑2
2
+
(ℎ1
2 − ℎ𝐿𝐿
2)
2𝑑𝑑2
𝐾𝐾2
By mass balance conservation: 𝑞𝑞𝐴𝐴(𝑑𝑑1) = 𝑞𝑞𝐵𝐵(𝑑𝑑1). From this, the value of ℎ1, the
hydraulic head value at 𝑥𝑥1 = 𝑑𝑑1 is:
Rechar9e N
iiiiiiiiiiiiiiiiiiiiiiiiiii
Water Tabfe
h
X
Hydraulic
conductivity K1
Hydraulic
conductivity K2 h1
unconfined Aquifer
lmpermea le Layer
L
Horizont al distance
Spring Emergence
---+ FlowqL
Channel
151
55
ℎ1 =
􀶩𝑁𝑁(𝑑𝑑1 + 𝑑𝑑2) +
𝐾𝐾1ℎ0
2
𝑑𝑑1
+
𝐾𝐾2ℎ𝐿𝐿
2
𝑑𝑑2
2𝑚𝑚
Where
𝑚𝑚 =
𝐾𝐾1
2𝑑𝑑1
+
𝐾𝐾2
2𝑑𝑑2
Hence, the specific discharge rate q at 𝑥𝑥1 = 𝑑𝑑1 + 𝑑𝑑2 is:
𝑞𝑞 =
𝑁𝑁𝑑𝑑2
2
+
(ℎ1
2 − ℎ𝐿𝐿
2)
2𝑑𝑑2
𝐾𝐾2
The dimensions and principal parameters are as used in Figures A3 and A4 above.
Here we reduce the permeability by two orders of magnitude for a peat layer of
1 m thickness, to a value of 0.043 m/day, typical of DHI’s estimates for a peat
soil. The Near Field simulation shows a 22% decrease in flow due to the 1 m peat
layer, and a total decrease of 24% if the water table is increeased by 0.5 m. The
Far Field simulation shows a 1% decrease due to the peat layer, and a 1.2%
decrease if in addition the water table rises by 0.5 m.
152
56
Figure A7. Schematic for Near Field flow path discharge calculation.
Figure A8. Schematic for Far Field flow path discharge calculation.
h0 = 415m
h0 = 550 m
Recharge = 24 mm/ year
m
K1 = 4.3d
lgnimbrite Aquifer
Unconfined Aquifer
lmpermea le Layer
L = 360m
hL = 400m
Recharge = 24 mm/year
i iiiiiiiiii i iii ii i i iiii ii il
m
K1 = 4.3d
lgnimbrite Aquifer
Unconfined Aquifer
lmpermea le Layer
L = 10,500m
m
K2 = 0.043d
Peat Layer
hL = 400 m
(+O.S m)
Channel
(+ O.Sm)
Channel
Floww/ opeat layer
(d2=0) and no water
table r ise:
73.0~
aay•m
Flow with peat layer
(d2=1 m) and no water
table r ise:
57.3~
......... day-m
Flow with peat layer
(d2=1 m) and 0.5 water
table rise:
55.4~
day·m
l'lq= 17.6 d:;,~m
24% decrease
Flow w/o peat layer
(d2=0) and no water
table r ise:
29.52~
day-m
Flow with peat layer
(d2=1 m} and no water
table rise:
-. 29.25--"c'_
____,,..- day•m
Flow with peat layer
(d2=1 m) and 0.5 water
table rise:
29.17~
day-m
l'lq =0.35~
day·m
1.2% decrease
153
57
Groundwater discharge
m3/m/day
NEAR FIELD
With channel in place 73.0
With 1m peat layer 57.3
% decrease in flow 21.6%
FAR FIELD
With channel in place 29.5
With 1 m peat layer 29.2
% decrease in flow 0.9%
RATIO of % CHANGE 23
Table A2. Summary of results – impacts of peat layer of 1 m thickness.
Groundwater discharge
m3/m/day
NEAR FIELD
With channel in place 73.0
With channel removed
0.5 m water table rise and
1 m peat layer
55.4
% decrease in flow 24.1%
FAR FIELD
With channel in place 29.5
With channel removed
0.5 m water table rise and
1 m peat layer
29.2
% decrease in flow 1.2%
RATIO of % CHANGE 20
Table A3. Summary of results – impacts of peat layer and channel removal/water
table rise.
154
58
3. Conclusions
While these calculations are simplified, they nevertheless show convincingly that
an inappropriate choice of a fixed water table elevation at the Near Field boundary
exaggerates the effects of water table rise and increased hydraulic resistance, by
the order of a factor of 20.
References
Dupuit, J., 1863. Etudes théorique et pratiques sur le movement des eaux dans les
canaux découverts at á travers les terrains perméables, 2nd ed., Dunod.
Forchheimer, P., 1886. Über die Ergiebigkeit von Brunnen-Anlagen und
Sickerschlitzen. Zeitschr. Archit. Ing. Ver. Hannover, 32, 539-563.
Verruijt, A., 1970. Theory of Groundwater Flow. Macmillan.
155
Expert Report
Peach, D.W. and Wheater, H.S., Concerning the Geology,
Hydrogeology and Hydrochemistry of the Silala River Basin

157
CONCERNING THE GEOLOGY, HYDROGEOLOGY AND
HYDROCHEMISTRY OF THE SILALA RIVER BASIN
Drs. Denis Peach and Howard Wheater
January 2019

159
iii
ABOUT THE AUTHORS
Dr. Denis Peach
Dr. Denis Peach spent nine years as the Manager of the British Geological Survey
(BGS) Groundwater Programme, then 6 years as BGS Chief Scientist. He is a
hydrogeologist with broad scientific interests and 46 years of experience which
includes work for a UK water authority, overseas work in tropical
hydrogeological environments, small island hydrogeology and work for
international consultants in arid zone hydrogeology. His particular scientific
interests include groundwater modelling which he developed in BGS, arid zone
hydrogeology, Chalk and Karst hydrogeology. He has been a Vice President of
the Geological Society of London (GSL), is a Visiting Professor at Imperial
College, London and University of Birmingham and he has given the prestigious
Ineson Distinguished lecture at the GSL. He has led geological and
hydrogeological research programmes in the UK and has sat on several national
research programme boards and national water resource strategic committees. He
currently carries out research with BGS and Imperial College and recently
provided advice to the University of Saskatchewan and UK engineering
consultants.
Dr. Howard Wheater
Dr. Howard Wheater is the Canada Excellence Research Chair Laureate in Water
Security at the University of Saskatchewan and was founding Director of the
Global Institute for Water Security. He also holds the posts of Distinguished
Research Fellow and Emeritus Professor of Hydrology at Imperial College
London. A leading expert in hydrological science and modelling, he has published
more than 200 refereed articles and 6 books. He is a Fellow of the Royal Society
160
iv
of Canada, the UK’s Royal Academy of Engineering and the American
Geophysical Union. He was awarded the 2018 International Prize for Water
(Dooge Medal) by UNESCO, the World Meteorological Organisation and the
International Association for Hydrological Sciences and the 2006 Prince Sultan
bin Abdulaziz International Prize for Water. He has initiated and led national and
international research programmes in the UK and Canada, and has advised states,
provinces and national governments on flood, water resource and water quality
issues. He was instructed by Hungary and Argentina at the International Court of
Justice, and recently sat on the Court of Arbitration concerning the Indus Waters
Treaty. He was, until 2014, vice-chair of the World Climate Research
Programme’s Global Energy and Water Cycle Exchange (GEWEX) project and
leads UNESCO’s GWADI arid zone water program. In Canada, he led the
Changing Cold Regions Network, focused on the analysis and prediction of
hydrological change in western Canada, and the Global Water Futures Program,
focused on managing water futures in Canada and other cold regions where global
warming is changing landscapes, ecosystems, and the water environment. His role
as Chair of the Council of Canadian Academies Expert Panel on Sustainable
Management of Water in the Agricultural Landscapes of Canada saw release of a
report in February 2013 entitled Water and Agriculture in Canada: Towards
Sustainable Management of Water Resources. In 2018 he was the only non-US
member of a US National Academies panel that reported on Future Water
Priorities for the Nation.
161
v
TABLE OF CONTENTS
1 INTRODUCTION.............................................................................................. 1
1.1. Experts’ Terms of Reference .................................................................... 1
1.2. Background to the report........................................................................... 2
1.3. Structure of the report ............................................................................... 3
1.4. Silala River basin – location and spatial extent ........................................ 4
2 SUMMARY OF FINDINGS.............................................................................. 6
3 GEOLOGY OF THE SILALA RIVER, RAVINE AND GROUNDWATER
CATCHMENT AREA ....................................................................................... 9
3.1. The geological context of the Silala River................................................ 9
3.2. Stratigraphy............................................................................................. 13
3.2.1. Stratigraphy of the Silala basin developed by Chile ................... 13
3.2.2. Stratigraphy developed by Bolivia.............................................. 17
3.2.3. Discussion ................................................................................... 21
3.3. Three-dimensional geology of the extended Silala groundwater
catchment.......................................................................................................... 25
3.4. Conclusions............................................................................................. 33
4 HYDROGEOCHEMISTRY OF THE SURFACE AND
GROUNDWATERS OF THE SILALA BASIN ............................................. 35
4.1 Introduction............................................................................................. 36
4.2 Discussion of chemistry analytical results .............................................. 38
4.3 Isotope analyses ...................................................................................... 40
162
vi
4.3.1. Interpretation of Oxygen-18 and Deuterium (δ18O and δ2H)
data ............................................................................................. 40
4.3.2. Carbon-14 and Carbon-13 data ................................................... 43
4.4 Conclusions concerning the origins of the spring waters in the Silala
groundwater catchment..................................................................................... 46
5 SUMMARY OF THE HYDROGEOLOGY OF THE SILALA
GROUNDWATER CATCHMENT - AREAS OF AGREEMENT AND
DISAGREEMENT BETWEEN CHILE AND BOLIVIA ............................... 47
6 DISCUSSION ON THE ENHANCEMENT OF SPRING FLOWS IN THE
CAJONES AND ORIENTALES WETLANDS BY THE USE OF
EXPLOSIVES .................................................................................................. 51
7 CONCLUSIONS .............................................................................................. 52
8 REFERENCES................................................................................................. 55
163
vii
LIST OF FIGURES
Figure 1-1. The Silala River and topographic catchment area, showing some of the
main physiographic features in and around the catchment. .................................... 5
Figure 3-1. Synthesis of geology for the region in which the Silala River basin is
located (SERNAGEOMIN (Chile), 2019). ........................................................... 11
Figure 3-2. Map showing a compilation and interpretation of the geology of the
Silala groundwater catchment (SERNAGEOMIN (Chile), 2019). ....................... 12
Figure 3-3. The updated integrated stratigraphic column of the Silala River basin
as mapped in Chile (SERNAGEOMIN (Chile), 2019). ........................................ 14
Figure 3-4. Stratigraphic column from SERGEOMIN Map 1 (SERGEOMIN
(Bolivia), 2017). .................................................................................................... 19
Figure 3-5. Disposition of Silala Ignimbrite (1.61 Ma) overlying Pliocene dacitic
lavas (2.6 Ma) (SERNAGEOMIN (Chile), 2019)................................................. 23
Figure 3-6. Schematic profile of Inacaliri-Apagado volcanic chain at the border of
Chile and Bolivia (SERNAGEOMIN (Chile), 2019)............................................ 24
Figure 3-7. DHI conceptual cross-section reproduced from DHI, 2018 (BCM,
Vol. 4, p. 88).......................................................................................................... 24
Figure 3-8. Geological cross section A-B from South West – to North East
through the Silala extended groundwater catchment showing the distribution of
lithological units and their stratigraphic positions and a cross-section C-D from
north west to south east through the Cajones and Orientales wetlands
(SERNAGEOMIN (Chile), 2019)......................................................................... 28
Figure 3-9. Amended map from DHI, 2018 (BCM, Vol. 4, p. 76, Figure 29)
showing in red (HGU 7) the DHI postulated fault system (SERNAGEOMIN
(Chile), 2019). ....................................................................................................... 30
Figure 3-10. Approximately horizontal jointing in the Silala Ignimbrite crossing
the Silala ravine with no displacement. Photo taken looking upstream at the
junction with the Quebrada Negra, (SERNAGEOMIN (Chile), 2019). ............... 31
Figure 3-11. Schematic structural profile in the SW sector of Silala River.
(SERNAGEOMIN (Chile), 2019)......................................................................... 33
Figure 4-1. Modified Stiff diagrams of the waters from Silala River area in Chile
and Bolivia (Herrera and Aravena, 2019b). .......................................................... 38
164
viii
Figure 4-2. Plot of δ18O and δ2H for river, spring water and wells water in the
rainy season (Herrera and Aravena, 2019a). ......................................................... 41
Figure 4-3. Plot of δ18O and δ2H for river, spring water and wells water in the dry
season (Herrera and Aravena, 2019a). .................................................................. 42
Figure 4-4. Distribution of 14C sampling points in the Silala River basin in Chile
(in the dry season) and Bolivia together with values of percent modern carbon
(14C pMC) (Herrera and Aravena, 2019a)............................................................. 44
LIST OF TABLES
Table 1. Compilation of the radiometric ages available from the Silala River area.
............................................................................................................................... 17
Table 2. Stratigraphic column for Volcanic rock units from SERGEOMIN
(Bolivia) in Figure 3-4 (SERGEOMIN (Bolivia), 2017, Map 1). Radiometric dates
are taken from Table 1. ......................................................................................... 20
165
ix
LIST OF ACRONYMS AND ABREVIATIONS
BCM - Bolivian Counter-Memorial
ca - About or approximately
CM - Chilean Memorial
cm - Centimetre
DHI - Danish Hydraulic Institute
GML - Global Meteoric Water Line
LML - Local Meteoric Water Line
l/s - Litres per second
Ma - Million years before present
SERGEOMIN
(Bolivia)
- Servicio Nacional de Geología y Minería de Bolivia
(Bolivian National Geology and Mining Service)
SERNAGEOMIN
(Chile)
- Servicio Nacional de Geología y Minería de Chile
(Chilean National Geology and Mining Service)
pMC - Percent modern carbon
μS/cm - Micro siemens per centimetre
13C - Carbon-13
14C - Carbon-14
2H - Deuterium
δ2H - Ratio of deuterium and hydrogen
18º - Oxygen-18
δ18O - Ratio of oxygen-18 (18O) and oxygen-16 (16O)

167
1
1 INTRODUCTION
1.1. Experts’ Terms of Reference
In the context of the dispute between the Republic of Chile and the Plurinational
State of Bolivia concerning the status and use of the waters of the Silala, to be
heard before the International Court of Justice, the Republic of Chile has
requested our independent expert opinion, as follows:
“Questions for Dr. Howard Wheater, as a hydrological engineer
(i) What are the major points of scientific agreement between
the Experts of Bolivia and those of Chile concerning the
hydrology of the Silala River?
(ii) What are the major points of scientific disagreement between
the Experts of Bolivia and those of Chile concerning the
hydrology of the Silala River?
(iii) What new evidence has been produced, since Chile submitted
its Memorial in July 2017, concerning the effect of the
channelization of the flow on Bolivian territory on the
watercourse of the Silala River that flows from Bolivia into
Chile?
Questions for Dr. Denis Peach, as a hydrogeologist
(i) What new evidence has been produced, since Chile submitted
its Memorial in July 2017, concerning the understanding of
the geology and hydrogeology of the Silala River?
(ii) Does the hydrogeological conceptual understanding and
parameterisation of the numerical models of Bolivia’s expert,
the Danish Hydraulic Institute (DHI), provide an adequate
168
2
basis to quantify the effects of channelization on the surface
water and groundwater flows from Bolivia to Chile?
(iii) Could the flow from groundwater fed springs in the Cajones
and Orientales springs have been significantly enhanced by
the use of explosives?”
In this report we consider the three questions to Peach. The questions to Wheater
are addressed in Wheater and Peach (2019). While this report represents our joint
opinion, the lead author has been Dr. Denis Peach.
1.2. Background to the report
In 2016 we were requested by the Republic of Chile to write two expert reports on
the Silala River, which were subsequently submitted to the International Court of
Justice in July 2017 as part of Chile’s Memorial (Wheater and Peach, 2017; Peach
and Wheater, 2017). At that time, the core of the dispute between Chile and
Bolivia was whether or not the Silala River is an international watercourse.
Following submission of the Bolivian Counter-Memorial (BCM) on 3 September
2018, we were requested to write expert reports to comment on the scientific
underpinning for the counter claims made by the Plurinational State of Bolivia.
We now understand that there is agreement between the parties on the central
point that the Silala River naturally flows from Bolivia to Chile. And as discussed
in Wheater and Peach (2019), there is also general agreement between Bolivia’s
and Chile’s experts about the nature and functioning of the natural hydrological
system.
The core of the remaining dispute between Chile and Bolivia is the quantitative
effect of the channelization of the Silala, in Bolivian territory, on the crossboundary
flow. The Bolivian expert consultant, Danish Hydraulic Institute (DHI),
169
3
estimates that the natural flows without drainage and channelization would be
30-40% less than the current situation (BCM, Vol. 2, p. 266). We disagree with
the DHI estimates of the effects of channelization. In our opinion the very large
estimates made by DHI are implausible.
DHI’s estimates are wholly based on the results of a DHI integrated groundwater
and surface water model, which purports to represent the natural hydrological
system that supplies spring flow to the headwaters of the Silala River. In this
report we examine the conceptual basis for the construction of the DHI numerical
model. We draw on further geological and hydrochemical investigations carried
out during 2018, subsequent to the submission of Chile’s Memorial (CM), to
improve our knowledge and understanding of the geology and hydrogeological
functioning of the Silala River and associated groundwater flows. We show below
that the DHI modelling is based on a flawed interpretation of the geology,
hydrogeological and hydrogeochemical functioning of the basin, and therefore has
no validity as a basis for detailed modelling of the effects of channelization on
surface water or groundwater flows.
1.3. Structure of the report
Section 1 describes the background to the report, its structure and the location of
the Silala River, ravine and catchment area. Section 2 answers the question posed
and briefly summarizes the major findings. Section 3 provides a description of the
geology of the Silala River basin, its ravine and groundwater catchment area, and
highlights the shortcomings of the DHI interpretations and evidence upon which
they have based their numerical model. Section 4 describes the hydrochemistry of
the surface and groundwaters of the Silala River basin and the origin of the
groundwaters and surface water of the basin and the significance of the origins for
groundwater modelling. Section 5 summarizes the hydrogeology of the Silala
170
4
River basin. Section 6 assesses the evidence for the enhancement of spring flows
by explosive methods as stated in the BCM. Section 7 draws some conclusions
and answers the questions for Dr. Denis Peach, as a hydrogeologist. Section 8 lists
the references cited in this report.
1.4. Silala River basin – location and spatial extent
The Silala River originates in Bolivia and flows into the Antofagasta Region of
Chile. It is one of the main tributaries of the San Pedro River. This, in turn, is a
tributary of the Loa River, the longest river in Chile (440 km long) and the main
watercourse in the Atacama Desert region, discharging into the Pacific Ocean.
More detail about the location is provided in the Chilean Memorial (CM, Vol. 1,
pp. 137-144).
Figure 1-1 shows the topographic catchment area of the Silala River and key
features of the river network. In CM, we noted the possibility of groundwater
recharge and inflows from areas beyond the topographic catchment, within
Bolivia (CM, Vol. 4, p. 273, Figure 7-1; Arcadis, 2017). Figure 1-1 shows an
estimate of this extended groundwater catchment, based on topographic and
geological analysis. This area is very similar to that identified by DHI as the
‘hydrological catchment’ (BCM, Vol. 2, Figure 5, p. 275) of the Silala River (with
minor differences due to the use of a different Digital Elevation Model), although
we, and DHI (BCM, Vol. 4, p. 103), acknowledge that it is possible that there may
be additional groundwater contributions from other, more distant, sources.
We note that the river originates in groundwater springs at the Cajones and
Orientales wetlands in Bolivia, which are the main source of its perennial flow at
the international border. The water supplying these springs is predominantly
derived from precipitation on the extended groundwater catchment area in
Bolivia, though the river also receives water from springs in Chile that are fed, at
171
5
least in part, from the topographic catchment area. The Chilean and Bolivian
experts agree that surface water runoff contributes a very minor proportion of the
average daily flow of the Silala River. In this report we discuss the geology of the
extended groundwater catchment, and hence the hydrogeological controls on
surface water and groundwater flows from Bolivia to Chile, informed by recent
analyses of water chemistry. In particular, we highlight important features of the
geology and geochemistry that Bolivia’s modelling fails to represent.
Figure 1-1. The Silala River (perennial drainage solid blue and ephemeral streams in
dotted blue lines) and topographic catchment area (outlined in black), showing some of
the main physiographic features in and around the catchment. The extended groundwater
catchment area for the Silala River is shown in green and includes a large area in
Bolivian territory.
Kilometers
MerotorPro;ection,WGSSol
Mllltary
0
.Po,t
.,,e,,toles
~ S/LALA RIVER BASIN
GROUNDWATER
CATCH/II ENT
--SILALA RIVER
TOPOGRAPHIC
CAT<:HMENT
~~ BRAZIL
PERU ( ~
'- ',
~-
BOLIVIA
.•·
>:
.".,'
u
0 PARAGUAY
u .;: L l, ARGENTINA
~
172
6
2 SUMMARY OF FINDINGS
In this section we briefly answer the questions posed in our terms of reference and
summarise our conclusions.
(i) What new evidence has been produced, since Chile submitted its Memorial
in July 2017, concerning the understanding of the geology and hydrogeology of
the Silala River?
New investigations in the Silala River topographic catchment have included field
observation, re-logging of borehole drill cuttings, geological mapping and
radiometric dating of the Chilean-named Silala Ignimbrite and Pliocene lavas.
This new information has revealed a more detailed understanding of the
stratigraphy in Chile and the extensive presence of a debris flow that lies at the
base of the Chilean-named Silala Ignimbrite and the upper boundary of the
Chilean-named Cabana Ignimbrite. It has also revealed a major fault in Chile,
a few hundred metres below the junction of the Silala River and the Quebrada
Negra tributary valley. The stratigraphy and this structure have not been
considered in the Bolivian hydrogeological conceptual understanding, or
incorporated into their numerical models. DHI introduce a new fault system
running through the Bolivian wetlands and down the Silala River ravine into Chile
(DHI, 2018), but no evidence to support this has been found in Chile.
New hydrogeochemical investigations have revealed the distinct character of the
spring and groundwater of the Quebrada Negra. And in conjunction with the
Chilean data, Bolivian chemical and isotopic analyses have revealed:
a) the distinctly different recharge origins of the spring water of the Bolivian
wetlands, Cajones (referred to in DHI, 2018 as the North Wetland or Bofedal) and
Orientales (referred to in DHI, 2018 as the South Wetland or Bofedal), and
b) the close similarities of the Chilean spring waters, recharged from perched
173
7
aquifers, to the spring and groundwaters of the Cajones wetland in Bolivia.
As with the geological structure and stratigraphy, above, this important difference
in recharge to the two Bolivian springs is not incorporated in DHI’s modelling.
(ii) Does the hydrogeological conceptual understanding and parameterisation
of the numerical models of Bolivia’s expert, the Danish Hydraulic Institute (DHI),
provide an adequate basis to quantify the effects of channelization on the surface
water and groundwater flows from Bolivia to Chile?
The DHI numerical models incorporate an incorrect stratigraphy and an
implausible fault system and take no account of the down-gradient Chilean
geological structure or the difference in origin of the Cajones and Orientales
spring waters. In particular,
a. The ignimbrite aquifer system identified in Chile (the Chilean Silala
and Cabana Ignimbrites), together with an interbedded fluvial debris
flow has not been recognized by DHI in their report (DHI, 2018), nor
incorporated into their models, neither has the vertical heterogeneity in
permeability. This will undoubtedly mean that the groundwater
flowpaths that they simulate as a result of their models’ permeability
distribution will be wrong.
b. The fault system that they propose will also affect the groundwater
flowpaths and the ease with which groundwater can move in their
invoked fault system region.
c. The different origins of the Cajones and Orientales spring waters are
due to the two distinct aquifer systems identified by Chile (with
considerable supporting evidence (sections 3, 4 and 5)), but these have
not been included in the DHI models.
174
8
d. The faulting mapped at outcrop downgradient in Chile and the
presence of Pliocene lavas (sections 3 and 5) between the two
(Chilean) ignimbrites (in Chile) which cause a decline in the
permeability of the Cabana and Silala Ignimbrites in Chile has
similarly not been considered.
We conclude that the DHI models do not simulate the groundwater system
properly and are unfit to quantify the effects of channelization in the Bolivian
wetlands or accurately represent the current hydrological system.
(iii) Could the flow from groundwater fed springs in the Cajones and
Orientales springs have been significantly enhanced by the use of explosives?
The evidence for showing that the groundwater-fed springs of the Cajones and
Orientales wetland has been enhanced by explosives is flimsy and the reference to
development of deep borehole yields by explosive methods is inapplicable. The
springs could not have been developed significantly to increase yields by the
explosive methods suggested by Bolivia.
In summary, we have shown that the numerical modelling results that have been
presented by Bolivia to demonstrate the alleged effects of channelization in the
Bolivian wetlands are incorrect. Their models are based on a misrepresentation of
the current hydrological system and the proposed scenarios. In short, with this
conceptual basis, their models could only produce implausible predictions.
175
9
3 GEOLOGY OF THE SILALA RIVER, RAVINE AND
GROUNDWATER CATCHMENT AREA
In this section, we discuss why Bolivia’s geological interpretation of the Silala
groundwater catchment is incorrect in several important respects. This
interpretation provides the basis for their understanding of the hydrogeology of
the catchment and hence for the construction of the numerical models developed
by the Bolivian experts, DHI (DHI, 2018), to simulate groundwater and surface
water flows within the Silala River basin. The geological interpretation presented
by DHI is implausible, and inconsistent with Chile’s data. Hence, we conclude
that DHI’s modelling is flawed and unsuitable as a basis for predicting the effects
of channelization.
3.1. The geological context of the Silala River
The regional scale geology in an area of approximately 20 km radius around the
Silala River is dominated by volcanic rocks. The outcrops of these rocks provide
evidence of the volcanic processes occurring in the region over the last
approximately 12 Ma (SERNAGEOMIN (Chile), 2017). Volcanism is often
episodic and between these events there may be periods of erosion and deposition
of sediments. Some of the oldest volcanic rocks that are exposed in this region
include sequences of ignimbrite rocks. These are permeable and form the major
aquifers in the region.
Covering the volcanic rocks are alluvial and colluvial sediments consisting of
sands, gravels including boulders and silts. These form minor local perched
aquifers in the Silala basin in Chile. Here they provide flows to many springs,
particularly along the northern side of the Silala River ravine.
Ignimbrites are deposited from explosive volcanic eruptions that extrude a mix of
volcanic gases, molten rock and ash in a highly fluid pyroclastic flow, flowing
176
10
under gravity at speeds of at least 100 km/hour (Wilson and Houghton, 2000).
Such flows are very destructive and tend to fill depressions and valleys in the
existing topography. They often travel long distances, up to tens of kilometres.
The other extrusive volcanic rocks that outcrop in the region, including the Silala
catchment, are lavas, which consist of more or less fluid molten rock. These lavas
erupted in a less explosive manner than the ignimbrites. They would have flowed
down-gradient from a volcanic vent, often at very low velocities, and travelled
much shorter distances. They can normally be seen around volcanoes, outcropping
radially around the volcanic centre, and often appear like lobes. These can be seen
on Figure 3-1 (e.g. PPlv – in purple and Msv – in pale brown) and Figure 3-2
(e.g.Pliv(a) – in purple and Msvd- in pale brown) and reflect the way the lava
flows have moved down slope from the eruption vent. The younger lava flows, if
andesitic or basaltic, are often permeable. Other rocks that outcrop high on the
sides of the volcanoes include glacial till, composed of rock fragments in a matrix
of clay. These were deposited by glaciers in the Pleistocene ice ages and normally
have low permeabilities. Fluvial debris flows are common in volcanic regions and
their sedimentary nature provides a high porosity in which to store groundwater
and can provide high intergranular permeability. Because of their different
permeabilities, a numerical model must be built on a correct interpretation of the
geological sequences.
As part of the studies and investigations referred to in section 1.2, further field
studies, geological mapping and rock dating have been carried out by the Chilean
National Geology and Mining Service (SERNAGEOMIN), since the submission
of Chile’s Memorial in September 2017. These are reported in SERNAGEOMIN
(Chile), 2019.
177
11
Figure 3-1. Synthesis of geology for the region in which the Silala River basin is located.
Solid black line corresponds to the topographic catchment area of the Silala River
(SERNAGEOMIN (Chile), 2019). The green line corresponds to the extended
groundwater catchment, as shown in Figure 3-2 (SERNAGEOMIN (Chile), 2019,
amended from SERNAGEOMIN (Chile), 2017).
The rocks that can be found outcropping in the Silala River catchment are shown
on the detailed geological map reproduced in Figure 3-2 (SERNAGEOMIN
(Chile), 2019). This map has been compiled from the recently updated studies of
SERNAGEOMIN (Chile), 2019; and from the report by SERGEOMIN (Bolivia),
2017. On Figures 3-1 and 3-2 the yellow colours represent ignimbrites and the
purple and light brown colours represent different ages of lavas.
D Msv MiouneVokank Raclts
D PPlv Plloctn+-Plelstocene
VOlcanlc Rocks
• · CJ P1iis Silala lgnimbrite
D Piic cabana lgnimbrite
· · · · · · D P1Hs Non-consolidate deposlu
SILALA RIVER
' . TOPOGRAPHIC
.. . CATCHMENT
I
. . I SILALA RIVER BASIN
"\-. GIOUNDWArER
CAlCHMENT ·· t /
178
12
Figure 3-2. Map showing a compilation and interpretation of the geology of the Silala
groundwater catchment (SERNAGEOMIN (Chile), 2019). This includes Bolivian territory
and used Bolivian maps and data from SERGEOMIN (Bolivia), 2003 and 2017. The map
shows the line, A-B, of the cross section on Figure 3-8 and the green line, C-D, of the
cross-section, also on Figure 3-8.
The compilation shown in Figure 3-2 is in many respects similar to the maps in
SERGEOMIN (Bolivia), 2017; but there are fundamental differences of
interpretation that significantly affect the hydrogeology of the area and the
construction of any numerical model. The first significant difference is in the
sequence of geological layers and their age relationships. This is known as the
stratigraphy, which is explained below in Section 3.2, and provides the
fundamental underpinning for the geological interpretation. In section 3.3 we
visualise the three-dimensional geology, from the edge of the Silala basin
groundwater catchment in Bolivia across the international border into Chile,
where the difference in Chilean and Bolivian stratigraphy is shown to lead to
Cl Hf Auvial deposits from the Holocene
D Ha Alluvial deposib from the Holocene
0 PIH(pc) Pyrodastic fall deposits
D PIHa Alluvial deposits from the UpperPlelstoc:ene-Holoc:ene
Cl Plslal Alluvial deposib from the Upper Pleistocene
c::J Pig Glacial deposits (Upper Pleistocene)
D Pliis Silala lgnimbrite (PltOcene-P~istoce ne)
~ Pllv(a) Vokank Sequenc:esfrom the lower Pleistocene (ca. 1.s Ma)
Cl Plivlbl Ryolithiclavadome{ca.1.SMal
D Psvd Vokanic Sequences from the Upper Pliocene (co. 2.6 Mal
C] Piic
0 Msvd
Cabana lgnimbrite [lower Pliocene; co. 4.12 Ma)
Vokank Sequenc:esfrom the Upper Mlotene tea. 6.6-5.8 Ma)
SILALA RIVER BASIN
GROUNDWATER
/ CATCHMENT
179
13
important differences in the geological structure that underpins the hydrogeology
of the basin. This misinterpretation of the stratigraphic succession leads to an
incorrect interpretation of the geology and consequently an incorrect
hydrogeological conceptual understanding and model.
There are also differences in interpretation regarding the faulting of the geological
sequence and, in particular, the existence of a fault system in the area of the
Bolivian wetlands and down the Silala River ravine from Bolivia into Chile,
proposed by Bolivia (Section 3.4). Large faults like this can provide high
permeability groundwater pathways or conversely may be barriers to transverse
groundwater flow.
As we discuss below in more detail, we conclude that DHI’s numerical models are
based on a misunderstanding of the hydrogeology, which will inevitably lead to
errors in model predictions.
Figure 3-2 also shows the outcrops of more recent volcanic deposits in Bolivia,
the presence of which is agreed by the parties. Overlying these are alluvial
deposits and glacial deposits (SERNAGEOMIN (Chile), 2017 and 2019; Arcadis,
2017). Within the alluvial deposits there are perched aquifers, which are important
in feeding springs along the Silala River in Chile. These also outcrop in Bolivia,
and hydrogeochemical evidence suggests they are important in supplying a
proportion of the spring waters that support the Bolivian Cajones and Orientales
wetlands.
3.2. Stratigraphy
3.2.1.Stratigraphy of the Silala basin developed by Chile
In Chile, five main bedrock geological units have now been recognized in the
Silala topographic catchment. Further radiometric age determinations of the rocks
180
14
in these units have been used to help construct the stratigraphy shown in
Figure 3-3. A compilation of the radiometric ages available is listed in Table 1.
Figure 3-3. The updated integrated stratigraphic column of the Silala River basin as
mapped in Chile (SERNAGEOMIN (Chile), 2019). The references for the radiometric
dates are found in Table 1. (Here the debris flow, see below, is identified as “Alluvial
Deposits from the Upper Pliocene-Lower Pleistocene”).
In summary the stratigraphic succession of volcanic rocks, and including a debris
flow, beginning with the oldest rocks outcropping in the Silala basin in Chile is as
follows:
Volcanic Sequences from the Upper Miocene-Pliocene (MsPvd) ca 6.6 – 5.8
Ma (see Table 1). These comprise a series of volcanic rocks including domes,
lava domes, lava flows and autoclastic breccia. This unit has been correlated with
the older parts of the Inacaliri and Apagado volcanoes, which have been
- I
A A A
Volcanic sequences · ····· A A A A
A A A of the Upper Pliocene A A A A
Psvd · ca. 2.6 Ma A A A
A A A A
A
Cabana lgnimbrite ··-- --
ca.4.12 Ma
I\ I\ I\
I\ I\ I\
I\ I\ I\
A
A
A
I\
I\ I\
I\
Volcanic Sequences from the Lower Pleistocene
Pliv •ca. 1.6·1 .1 Ma
•• T2 ... ....... ... Terrace II: 8,430 cal yrs BP
(Latorre & Frugone, 2017)
n .......... Terrace I: 530·670 cal yrs BP
(Latorre & Frugone, 2017)
,-
·········· Silala lgnimbrite
Pliis·ca.1.61 Ma
········· Alluvial Deposits
from the Upper Pliocene · Lower Pleistocene
······· Volcanic Sequences I\
from the Upper Miocene · Pliocene I\ I\
MsPvd · ca. 6.6 · 5.8 Ma
181
15
radiometrically dated at 5.8 Ma. They are less fractured than the Ignimbrites and
are likely to be considerably less permeable.
Cabana Ignimbrite (Piic) ca 4.12 Ma (see Table 1). The Cabana Ignimbrite is a
medium to poorly welded tuff of white and white-pinkish color with vesicular and
dacitic pumice and subangular and angular lithics in an ash matrix. In Chile
(SERNAGEOMIN (Chile), 2019) in borehole CW-BO at the international border,
it was found to be over 53 metres thick. Recently an age date of 4.12 Ma
(SERNAGEOMIN (Chile), 2017) has been determined. The Cabana Ignimbrite
has been found, in an overflowing borehole (SPW-DQN) and in boreholes drilled
in 2016, to support considerable groundwater flows (Arcadis, 2017).
Debris flow (date by stratigraphic position). A thin (20 cm) fluvial deposit can
be seen at outcrop near the Inacaliri police station in Chile (SERNAGEOMIN
(Chile), 2019) directly overlying the Cabana Ignimbrite and underlying the Silala
Ignimbrite (see below). Borehole CW-BO, drilled very close to the international
border was cored and a debris flow identified overlying the Cabana Ignimbrite
(Arcadis, 2017; SERNAGEOMIN (Chile), 2017 and 2019). This was 13 metres
thick but has been correlated with the fluvial deposits seen at outcrop, because a
debris flow is a fluvial deposit and would be expected to become finer grained as
it flows down gradient, while still occupying the same stratigraphic position in the
geological succession. In a pumping test (Arcadis, 2017), the upper strata of the
Cabana Ignimbrite and the debris flow were found to support groundwater flow.
Volcanic sequences from the Upper Pliocene (Psvd) ca 2.6 Ma (see Table 1).
These dacitic lavas are of pale grey color and can be found at outcrop beneath the
Silala Ignimbrite (1.61 Ma (see below)). They have been dated at 2.6 Ma and have
also been found, in borehole MW-DQN, lying beneath the Silala Ignimbrite
(SERNAGEOMIN (Chile), 2019). These lavas have low levels of fracturing and
are likely to have low hydraulic conductivity.
182
16
Silala Ignimbrite (Pliis) ca 1.61 M (see Table 1). This ignimbrite is a welded
tuff of andesitic composition. It is pink and has distinct cooling units or flow
levels which can be seen dipping gently to the south west in the walls of the Silala
River ravine in Chile. In SERNAGEOMIN (Chile), 2017, the age of deposits was
bracketed at between 2.6 Ma and 1.48 Ma, since it overlies the dacitic lavas
(2.6 Ma, see above) and is covered by an andesitic lava flow (Volcanic sequences
of the Lower Pleistocene - 1.48 Ma, see Table 1) from the Inacaliri volcano in
Bolivia. A new radiometric date of 1.61 Ma (see Table 1) confirms its
stratigraphic position.
Pyroclastic Fall Deposits (PlH(pc)) ca 630 ka (see Table 1). These deposits
comprise well-stratified fine to medium-grained ash found in the central and
southern parts of the Chilean study area. Recently an age of 630 ka (see Table 1)
has been determined for the ash deposits. These deposits form a thin capping to
areas in the south of the topographic catchment and have a low hydraulic
conductivity. Infiltration tests (Arcadis, 2017) gave a low infiltration capacity.
Age (Ma) Unit Reference
630±310 ka Pyroclastic fall deposit
Blanco and Polanco,
2018
1.48 ± 0.02* Volcanic sequences of the Lower Pleistocene
Almendras et al.,
2002
1.612±0.018 Volcanic sequences of the Lower Pleistocene
Sellés and Gardeweg,
2017
1.61±0.08 Silala Ignimbrite (Chile)
Blanco and Polanco,
2018
1.74±0.02* Nlsg-Volcanic sequences of the Lower Pleistocene
SERGEOMIN
(Bolivia), 2003
183
17
Age (Ma) Unit Reference
2.6±0.4 Volcanic sequences of the Upper Pliocene
SERNAGEOMIN
(Chile), 2017
3.2±0.4* Ntpg-Ignimbritas Silala (Bolivian)
SERGEOMIN
(Bolivia), 2017
4.12±0.08 Cabana Ignimbrite (Chile)
SERNAGEOMIN
(Chile), 2017
5.84±0.09* MPv2-Volcanic sequences of the Upper Miocene
Almendras et al.,
2002
5.8±0.4* MPv2-Volcanic sequences of the Upper Miocene
Almendras et al.,
2002
6.04±0.07* Volcanic sequences of the Upper Miocene
SERGEOMIN
(Bolivia), 2003
6.63±0.06 Volcanic sequences of the Upper Miocene
Blanco and Polanco,
2018
6.6±0.5* Nis-3-Silala Ignimbrites (Bolivian)
SERGEOMIN
(Bolivia), 2017
7.8±0.3* MPvl-Silala Ignimbrites (Bolivian Nis 1) Ríos et al., 1997
Table 1. Compilation of the radiometric ages available from the Silala River area.
* Indicates a Bolivian radiometric date (SERGEOMIN (Bolivia), 2017), all other dates
detailed in SERNAGEOMIN (Chile), 2019.
3.2.2. Stratigraphy developed by Bolivia
The details of the stratigraphy developed by SERGEOMIN (Bolivia) in 2017 are
described below for comparison with the stratigraphy determined by geological
mapping and analysis in Chile. The stratigraphic column shown in Figure 3-4 and
184
18
Table 2 is from SERGEOMIN (Bolivia), 2017, Annex A, Map area 1. Map area 1
is highlighted since it pertains to the DHI near field model (DHI, 2018).
SERGEOMIN (Bolivia) have identified three Ignimbrite deposits (labelled Nis 1,
Nis 2 and Nis 3), and two debris flows, one occurring at the base of Nis 1 and the
other between Nis 2 and Nis 3. It should be noted that SERGEOMIN (Bolivia)
and DHI in their reports refer to all three ignimbrites in Bolivia as Silala
Ignimbrite. This is potentially confusing, because SERNAGEOMIN (Chile) refer
to two ignimbrites in Chile, the upper of which they have named the Silala
Ignimbrite and the lower the Cabana Ignimbrite, as detailed in Section 3.2.1. DHI
in their report (DHI, 2018) divide the ignimbrites into different lithological types
depending upon whether they are highly welded or less welded, but these
currently cannot be correlated directly with the Chilean divisions of Silala and
Cabana Ignimbrites.
Bolivia reports three dates for ignimbrite rocks in the Silala extended groundwater
basin, in Bolivia. These are 7.8 Ma, 6.6 Ma and 3.2 Ma (see Table 1). The oldest
of these dates (7.8 Ma) is stated to be from the base of an outcrop of Bolivian
Silala Ignimbrite Nis 1 in the Silala ravine (in Bolivia), the 6.6 Ma date is from
rocks much further to the north and stated to be from Bolivian Silala Ignimbrite
Nis 3. The 3.2 Ma date is from ignimbrite located towards the northern edge of
the extended groundwater catchment and has not been correlated with the
Bolivian Silala Ignimbrites (Nis 1, Nis 2 and Nis 3) (SERGEOMIN (Bolivia),
2017). For comparison, the date for the Chilean Silala Ignimbrite is 1.61 Ma and
the date for the Chilean Cabana Ignimbrite is 4.12 Ma.
185
19
Figure 3-4. Stratigraphic column from SERGEOMIN Map 1
(SERGEOMIN (Bolivia), 2017).
u
C • l
N
0
E ~
C A
•
N
N
~'
0
C
0
STRA11G!RAPHIC COWMN
CHARACTERIS,TilCS OF GEO'LOG'ICA!L UNITS
Sedim entai;y units
~ AJ DEp:i&t
~ c«u-1~ 1EJ!01S11 Gra.>=I. Q111. an aM dayS
~ CdU'o1al depo:;ll
~ iill/llo-gacial ~ Qa-"21, ,samll amll a ;;,~
V,o!leani,c units
\fot..CAN IC- a!Al!I
- A. l!E p RI)
2
as Sllala . ra (lO;!
as Sllala Ctiloo
2
~tl)l:f.k: lllaJit "11l1EI. ~ ,ar tl lll!DS.
ba!lilog.Aooesf.:6 ,Cf ~j(, pL
~Jc lllaJit gra)'. Hlitlated ortl lll!DS.
ba!lilog. ,"-l!lle5ttes (6 hb. ,lllg-bl-hb
Pc."]ltrjl:t'Je ~ !IJa'J'', lan:ln.r._ ,or rn lll:lltE.. new
taming. ,"-l!lfe5ltEI; · p,:,c, C{g-!ll(-qix
Pc."]ltrjl:t'.le lllaJit gra)'. Hlitlated orti lllllCl;s. 11aW
taming. ,"-l!lfe5ltEI; 11b. .ad&tlt-ffll
~11'.le gra)'. lamlna.te:I ortl lllll'Cts.
fllllll~Mi:lesle ci' tit .1i:IHJl-lit;.qlX
Dad!IC-t.J llml .ea t.. ~~ sl obs,, roltth
11;1;acpe11 ~ ~
B«Mn gra'}' dEjXiS,'ts al jlEtljl:6. ~
jlllil(=taile-. =11;.
Dadtlc.~ntrb:t'.: llllWlll IJOJ!trpr.'.C. I l:t'.ooiE.
.illllle C mooootdlc.,Cla9::c ~ . ~ ·l!il'
Dadllc. ig111RH!te llmlll1 ~ui:. 11f11 jllnk.
ti IIIOCU, 'i\1111 .:mlE!'.lale· f'IJI, qz:~;j)
Glil)' IH0',111 ~ i'eliilMffi, QTai
\l:llra:JIC gla85. ptm:ca;tooe. 5aild6.
186
20
Map
Abbreviation Age dates Volcanic Unit Lithology
Qlin2 1.48 Ma Lavas Inacaliri 2 Andesite
Nlcn 1.74 Ma Lavas Cerro
Negro Andesite
Nllsg Lavas Silala
Grande Andesite
Nlin1 5.84 Ma Lavas Inacaliri 1 Andesite
Nlsc Lavas Silala
Chico Andesite
Nis-3 6.6 Ma Ignimbrite Silala 3 Ignimbrite
Nfd-2 Deposit flow 2 Pebbles, gravel, pumice,
sand
Nis-2 Ignimbrite Silala 2 Ignimbrite
Nis-1 7.8 Ma Ignimbrite Silala 1 Ignimbrite
Nfd1 Deposit flow 1 Pebbles, gravel, pumice,
sand, volcanic glass
Table 2. Stratigraphic column for Volcanic rock units from SERGEOMIN (Bolivia)
in Figure 3-4 (SERGEOMIN (Bolivia), 2017, Map 1). Radiometric dates are
taken from Table 1.
SERGEOMIN (Bolivia), 2017, identify lavas from the Cerrito Silala (Chico in
Bolivia) and the early lavas from Volcán Inacaliri and lavas from Volcán
Apagado (Silala Grande in Bolivia). All of these rocks date from approximately
6.6 – 5.8 Ma (see Table 1). They have taken the earlier two radiometric dates for
the Ignimbrites (Nis 1 and 3) to place them stratigraphically below the lavas of the
Cerrito Silala and the earliest lavas of the Volcán Inacaliri and Volcán Apagado.
This is in conflict with the interpretations of Chile. Based on the Chilean
evidence, the Silala Ignimbrite (in Chile) and the Cabana Ignimbrite are younger
than the lavas from Cerrito Silala (Chico in Bolivia) and lie stratigraphically
above these lavas.
187
21
There appear to be no lavas outcropping in Bolivia in the extended groundwater
basin with ages comparable with the Volcanic Sequences of the Upper Pliocene
(dated 2.6 Ma, see Table 1). However, SERGEOMIN (Bolivia) and DHI agree
with Chile that the more recent lavas from Inacaliri and Volcán Apagado of
Pleistocene age overlie all the above-mentioned older volcanic rocks (see
Figure 3-2).
3.2.3. Discussion
The main differences in geological interpretation between Bolivia and Chile are
focused on the ignimbrite deposits. It is clear that there is a thick succession of
separate ignimbrite deposits that have filled the Silala basin in Bolivia and that
only some of these outcrop in Chile. Ignimbrites with ages as found in Chile, of
1.61 Ma and 4.12 Ma, have not been found in Bolivia. There are five age dates for
the Ignimbrites (see Table 1), which represent separate volcanic events, over a
large time range:
a. 7.8 Ma – Bolivian Silala Ignimbrite Nis 1
b. 6.6 Ma – Bolivian Silala Ignimbrite Nis 3
c. 4.12 Ma – Chilean Cabana Ignimbrite
d. 3.2 Ma – Bolivian Silala Ignimbrite (Ntpg)
e. 1.61 Ma – Chilean Silala Ignimbrite
Also, there are identified debris flows that lie between various ignimbrite
deposits, one in Chile and two in Bolivia, but these do not appear to be correlated.
However, they are important, because in Chile there is drilling evidence and
pumping test results that show that they contribute to strong groundwater flow
horizons (Arcadis, 2017). There is more or less agreement between the parties on
the stratigraphy of the Pleistocene and more recent deposits.
188
22
It would seem highly unlikely that the Silala (Chilean) Ignimbrite outcropping in
Chile with a well-defined, recently analyzed, radiometric date of 1.61 Ma, which
is continuously traceable in the Silala ravine walls in Chile to the international
border, should, further upstream in the ravine in Bolivia, have an age of 7.8 Ma.
This is implausible and would require a large geological structure such as a fault
with a throw of many tens if not hundreds of metres, for which no evidence has
been presented by SERGEOMIN (Bolivia) or DHI (see section 3.4). The age of
the Chilean-named Silala Ignimbrite is well defined by several radiometric dates
from underlying deposits. The lavas of the Volcanic Sequences of the Upper
Pliocene, which can be seen in outcrop underlying the Silala Ignimbrite, as shown
on Figure 3-5, have been dated recently (Table 1) at 2.6 Ma. Also, the underlying
Cabana Ignimbrite (in Chile), found in several boreholes drilled in 2016 (Arcadis,
2017), has a recent radiometric age date of 4.12 Ma, all considerably younger than
7.8 Ma. The outcrop of the Silala Ignimbrite (named in Chile) clearly crosses the
international border, and more generally there is very strong evidence that the
Ignimbrite succession in Chile also occurs across the border in Bolivia. The
Chilean ignimbrite dates confirm that, in Chile, the oldest rocks are the Volcanic
Sequences of Miocene/Pliocene, which underly the Cabana Ignimbrite. If, as
dated in Bolivia, there are older Ignimbrites outcropping in the basin, these must
underly the younger Miocene/Pliocene lavas, but the higher and younger
ignimbrites (in Chile, Silala and Cabana Ignimbrites) overlie these lavas. For
clarity, a conceptual cross-section at the international border has been constructed
on the basis of the radiometric dates and outcrop relationships discussed above, is
shown in Figure 3-6.
For comparison, Bolivia’s conceptual cross-section through the geology at the
International border is reproduced from DHI, 2018, in Figure 3-7. The differences
in interpretation of the geology can be clearly seen. On DHI’s cross-section a fault
zone is depicted (see also section 3.3) but there appears to be no displacement of
189
23
the rocks either side. Geological faults occur when great pressures built up in the
earth’s crustal rocks are relieved, and this results in movements of the rocks on
either side of a fault plane. Movements can be as small as a few centimeters or
very large, and they might be horizontal movements or vertical movements or low
angle movements. Clearly when the rocks move they are split along the fault
plane and a displacement occurs. On the DHI conceptual diagram (Figure 3-7), a
horizon of welded ignimbrite is shown crossing the ravine. Even though a major
fault zone is depicted on the Figure, the welded ignimbrite is shown at the same
level in both walls of the ravine, implying that the DHI-inferred major fault zone
has caused no displacement. This is so unlikely that we believe it impossible.
Figure 3-5. Disposition of Silala Ignimbrite (1.61 Ma) overlying Pliocene dacitic lavas
(2.6 Ma) (SERNAGEOMIN (Chile), 2019).
190
24
Figure 3-6. Schematic profile of Inacaliri-Apagado volcanic chain at the border of Chile
and Bolivia, including Cerrito Silala, Cerro Inacaliri and Volcán Apagado, showing the
Silala Ignimbrite underlain by the Cabana Ignimbrite, which is in turn underlain by
Miocene/Pliocene lavas of Cerrito Silala, Volcán Apagado and the Volcán Inacaliri.
Beneath all are earlier Ignimbrite deposits. The purple shows the Pleistocene lavas of
Inacaliri and Apagado, (SERNAGEOMIN (Chile), 2019).
Figure 3-7. DHI conceptual cross-section reproduced from DHI, 2018
(BCM, Vol. 4, p. 88, Figure 36).
NW
Volcan l nQllrt
Cerro lnacahri
LEGEND
~ lJ11dirf!"li-t ·li,.lrtl l,.,~i-•t< 1
~ f).,( iii: ltif lhl - hh}: C'.rixmo 18f1;m/N,lr
c=J Ocon: and andesltlc avas
Surface Water Inflow
- AlWaldeposts
Regional
Groundwater 7
Surfoa, and Ground
K 0.5 m/d
Volc!in Apagado SE
............... ,. ..... ., :, ......
--- ......: _ .............. ......
,.- -.... " - ~I\_"
" :_ ~ _"_ ~ ,.
" A A II A A
" " " " ,.
Gloc1er
Groundwate:r
- Bofedel (Organic Matter)
CJ Channel Alluvium
~ Alluvium / Colluvlum
i::J Loe.al Andeslllc Volcanism
- Loal Oadtlc Volcanism
- RegJonal lgnlmbflte
Volcan Apiigo<lo
K < OSITl/d
Welded and Fradured
lgnlmbnte
191
25
3.3. Three-dimensional geology of the extended Silala groundwater
catchment
In order to visualize the geology of the extended groundwater catchment so that a
good understanding of both the regional and local hydrogeology might be gained,
the map shown in Figure 3-2 was compiled and the profile A-B (cross section) in
Figure 3-8 was constructed. This was achieved by studying the geological maps
and reports of SERGEOMIN (Bolivia), 2003 and 2017; SERNAGEOMIN
(Chile), 2017 and 2019, including all the radiometric ages available (see Table 1),
and satellite images from Google Earth. The map (Figure 3-2) is very similar to
the version of SERGEOMIN (Bolivia), 2003. This compilation has been enabled
after the submission of the BCM and the associated DHI, 2018 report, which
referred to various SERGEOMIN (Bolivia) reports (2003 and 2017). These were
subsequently requested from, and provided by, Bolivia. Field studies and
observations in Bolivia, further petrographic study and radiometric dating have
not been possible.
Volcanic centres, including Cerrito Silala, align in an approximate North-South
direction (SERNAGEOMIN (Chile), 2017 and 2019). These volcanic centres are
aligned in this way in response to a local crustal extension (i.e. pulling apart of the
earth’s crust) which produced a plane of weakness that upwelling magma (molten
rock) under pressure took advantage of, allowing the formation of a line of
volcanoes (volcanic centres). These centres have ages of 6.6-6.0 Ma (see Table 1).
The volcanoes of Inacaliri and Apagado form a topographic high, which together
with the north-south alignment mentioned above have meant that the pyroclastic
flows that deposited the Silala and Cabana Ignimbrites in Chile had a topographic
high to overtop in order to flow downgradient to the south west into Chile.
The geological cross section A-B, shown in Figure 3-8, visualizes the geology
with depth. The line of the section (see Figure 3-2) goes from the edge of the
192
26
groundwater catchment in Bolivia to the south west, approximately along the line
of the Silala River. The paucity of borehole information limits the threedimensional
accuracy of geological knowledge and understanding. Nevertheless,
the compilation of Chilean and Bolivian data (radiometric dates in Table 1, field
observations that where available in Bolivian reports, and the Chilean mapping
observations) is expected to give the best understanding to date of the geology of
the extended catchment.
If the Bolivian 7.8 Ma ignimbrite age date is valid then there must be ignimbrite
deposits beneath the Miocene/Pliocene lavas of Cerrito Silala, since these are
younger, and this has been shown on Figure 3-8. The Chilean evidence shows that
there are at least two Ignimbrite deposits overlying the Miocene/Pliocene lavas in
Chile. Since the Bolivian stratigraphy does not include ignimbrite deposits
overlying the Miocene/Pliocene lavas, it is incorrect. The Bolivian geological
succession beneath the Pleistocene lavas of the Inacaliri and Apagado volcanos
cannot be correlated with the Chilean geological succession.
A cross-section through the Cajones and Orientales wetland areas, Figure 3-8,
shows Chile’s geological interpretation with depth. This clearly shows the
Chilean-named Silala and Cabana Ignimbrites overlying the Miocene/Pliocene
volcanics of the Cerro Silala (Chico in Bolivia). This three-dimensional
geological configuration has not been used by DHI in the construction of their
models. Instead they use a stratigraphy in their models that results in different
hydrogeological layers and they invoke a fault system which perhaps justifies
their hydraulic parameter distribution. The restricted region of Chilean Silala and
Cabana ignimbrite, through which groundwater must flow beneath the Silala
River, can also be seen in Figure 3-8.
193
Figure 3-8. Geological cross section A-B from South West – to North East through the Silala extended groundwater catchment
showing the distribution of lithological units and their stratigraphic positions (for legend see Figure 3-2) and a cross-section C-D
from north west to south east through the Cajones and Orientales wetlands. The black dashed vertical line represents the North –
South volcanic centre alignment through Cerrito Silala (SERNAGEOMIN (Chile), 2019).
27
SW
m as/
5,000
PIHa
4,000 I
3,000
Hf
I
PIH(pc)
Si/ala
River
PIH(pc)
Piic
PIH(pc) Psvd
Si/ala
River MWS-UQN
PW-UQN
MWL-UQN
SPW-DQN Pls(a)
I
CHILE I BOLIVIA
MW-B01
PW-BO I
CW-BO
Si/ala I
River I
I
I
I
Si/ala
River
Pliv(a) Piic
MsPvd
Pls(a)
~~~~~~~~~~% _________________________________ _ - - r - - - - - - - - - - - _______ _ _______________________________ _
,- ,- ,- ,- ,- ,- ,-,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- r- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,- ,-. r.e . .. .. .......... ""." ......... . " .................... " .. ,.... ,.... ,- ,- ,.... ,- ,- ,- ,- ,- ,- ,- ,.... ,- ,- ,- ,- ,- ,- ,- ,- ,- ,.... ,- ,.... ,- ,- ,- ,- ,- ,- ,- ,- ,.... ,- ,- ,- ,- ,- ,- ,.... ,- ,.... ,- ,- ,- ,- ,- ,- ,- " .. "."............................ . . . . . . . . . . . . .
- - - ;-_-_ c;-: : ~ ::_:_:_~-:-:-:-:-:::-:-:-:-:-:-:--------- --~----8:;-----~-----~-----=-z--:--~-----~----~-----:-:-:-2--~-----~-----z=::----==:::---- :s,:-:___:__-:----=----=----=----=-----=-----=--~---=--=----=- --=----=---=- ..:--~-;:~;: _c;:_::--_;:: - --;::-;:-:_ ~--~--~,-- ~- -~~--~- -- '_: !--~:-~-~-~--=--~-_:-::_- 7--
-.L: - - - • - - - - - - - - - - - - ' · - ~-~-- - --- -~ --- -=--=-.::- -=-r - r r - - - --r------- -- _cc_" ~ ~~ r - - - - - - - - - r - -
7.8-6.6 Ma lgnimbrites
NW e~---------------------------------------46) SE
mas/
4,440
4.420
4,400
4,380
4,360
Cerrito Silala
Msvd
A1l,' A
Piis
A Al A A A A r-
A A A A I" A A A A A A A A A A r- r-
A A /\ IA A A A A A A>---;c::~===:===:::;:::=-::::;--
Piis
A A Al A A A A A A A ---i --, --,
A A 1
1
/\ A A A A A A A A A A A A
/\AAAAAAAAAAAA
Los Orientales
Piic
4.440
Piis
PIHa
4.420
4.400
--, --, --, --, --,
4,380
4,360
NE
m as/
5,000
4,000
3,000

195
29
The DHI Near Field model does not incorporate the Chilean Silala and Cabana
Ignimbrite with the interbedded debris flow. These deposits, which form the main
deep aquifer in Chile, overlie the Miocene/Pliocene lavas, and must continue to
the north east over the International border into Bolivia. The DHI model
representation of the geology and hydrogeology does not incorporate this
geological configuration, proven in Chile, and is therefore inherently flawed.
There is agreement between the parties that the major regional aquifer is formed
by the ignimbrites, but in Chile these are underlain by much less permeable
Miocene/Pliocene lavas. Therefore, the groundwater flow path in the Silala and
Cabana Ignimbrites through to Chile is restricted, although both these deposits are
known to support high groundwater flows (Arcadis, 2017).
Another facet of the three-dimensional geology that can impact significantly on
the groundwater flow regime is the presence of geological faults. Geological
faults cause displacement of rock sequences, bringing strata of different nature to
lie next to each other. They can cause extensive fracturing or grind up the rocks to
a fine powder, which can line the fault planes, producing a region of low
permeability. Thus, they can form high permeability pathways for groundwater or
low permeability barriers to groundwater flow.
The modelling that was carried out in support of the BCM (DHI, 2018) has
employed high hydraulic conductivities along an alleged fault zone, which is
shown on their maps (SERGEOMIN (Bolivia), 2017) as running from the
Orientales wetland to the Cajones wetland and bending around to follow the line
of the Silala River to cross the international border into Chile (see Figure 3-9).
However, no evidence, including displacements, fault gouge deposits or rock
shattering has been found in Chile to support the presence of such a fault. No
evidence of large displacement is provided by SERGEOMIN (Bolivia) in their
2003 or 2017 reports and DHI, in their conceptual cross-section (Figure 3-7),
show no displacement across the fault. SERGEOMIN (Bolivia), 2017 provide
196
30
evidence of fractures and their directions and some minor faulting with minor
displacements in Bolivia, but nothing of such a major character as the major fault
system introduced by DHI.
Figure 3-9. Amended map from DHI, 2018 (BCM, Vol. 4, p. 76, Figure 29) showing in
red (HGU7) the DHI postulated fault system (SERNAGEOMIN (Chile), 2019).
This fault system has been assumed to be vertical by DHI, 2018, yet such a fault
system showing such outcrop sinuosity could only geometrically occur in the
manner assumed by DHI, 2018, if it had a very low angle. DHI specify this fault
system to be 400 metres deep and 50 metres across and provide it an elevated
-- Main canals
- HGU7
D HGU8
D HGUl
D HGU2
D HGU3
D HGU4
- HGU5
D HGU6
CHILE
Ara11
Enlar
BOLIVIA
197
31
hydraulic conductivity down to 200 metres depth (DHI, 2018). On the geological
maps provided with SERGEOMIN (Bolivia), 2017, the faults appear as inferred.
However, they have not been found at outcrop in Bolivia by SERGEOMIN
(Bolivia), 2017, or by SERNAGEOMIN (Chile), 2019, in Chile. Furthermore, the
morphology of the walls of the Silala ravine at the international border and
downstream of the border which are composed of Silala (Chilean notation)
Ignimbrite shows no signs of displacement, and major joints can be seen at
approximately the same level on each side of the ravine (see Figure 3-10), clearly
continuous across the ravine.
Figure 3-10. Approximately horizontal jointing in the Silala Ignimbrite crossing the
Silala ravine with no displacement. Photo taken looking upstream at the junction
with the Quebrada Negra (SERNAGEOMIN (Chile), 2019).
198
32
In short, there appears no evidence for this fault system introduced by DHI, yet it
is a dominant feature of the hydrogeology used by DHI to model the Near Field
system.
A vertical normal fault has been mapped in Chile (SERNAGEOMIN (Chile),
2019), trending N-S, which affects the front of the dacitic lava flow dated 2.6 Ma
(Table 1, Volcanic sequences of the Upper Pliocene) but does not displace the
overlying Chilean Silala Ignimbrite. This tectonic event (fault) occurred between
2.6 to 1.6 Ma (see Figure 3-11). This structural configuration and the presence of
the low permeability Pliocene lavas under which the Cabana Ignimbrite occurs,
and the thinning of the Silala Ignimbrite to only the upper 8 metres
(SERNAGEOMIN (Chile), 2019), causes a reduction in the transmissivity of the
Ignimbrite aquifer. As a consequence, an elevated groundwater piezometric
surface is observed at the borehole SPW-DQN (Suárez et al., 2017) from which
groundwater overflows to the Silala River at rate of approximately 90 l/s
(Suárez et al., 2017). The overflow began during the drilling and did not begin
until the borehole depth was 28 metres (pers com. Muñoz, 2017). This clearly
indicates a considerable variation of permeability with depth, since here the upper
several metres of the Chile-named Silala Ignimbrite act as a low permeability
layer that confines groundwater in the lower layers and in the Cabana Ignimbrite,
which is found at depth (SERNAGEOMIN (Chile), 2019). The groundwater flows
down-gradient to the south west are very much reduced due to the existence of the
fault, the presence of the Pliocene lavas beneath the Chilean Silala Ignimbrite and
the underlying Cabana Ignimbrite being at greater depth. In consequence, the
groundwater levels in the Cabana Ignimbrite are very low, as evidenced in
borehole EW-PS (SERNAGEOMIN (Chile), 2019; Arcadis, 2017). Hence, there
are important geological features in Chile that significantly affect groundwater
flows across the border, but these are not recognized by Bolivia, or included in
DHI’s modelling.
199
33
Figure 3-11. Schematic structural profile in the SW sector of Silala River.
(SERNAGEOMIN (Chile), 2019).
3.4. Conclusions
The geology of the extended groundwater catchment of the Silala River is highly
complex. Clearly there has been a succession of volcanic pyroclastic flow events
over several million years with intervening erosive periods during which
sediments in the form of debris flows and minor fluvial deposits have been laid
down. The thickness of these ignimbrite deposits is unknown but likely to be
large, perhaps over 200 metres, but this has not been proven by drilling.
There are four main differences in the Chilean interpretation of the geology to that
in the DHI report (DHI, 2018), which have impacts on the hydrogeological
conceptual model that DHI, in their models, are attempting to represent
numerically.
SW
MsPvd
Pliis
N-S normal fault system (2.6-1.6 Ma),
morphologically is a "soft" flexure
NE
Dacitic lava flow
(2.6 Ma)
A A 'i A •
MsPvd
200
34
The first is the stratigraphic position of the two ignimbrite deposits recognized
along the Silala River in Chile. The contact relationships observed in Chile, as
discussed above, and the ages obtained (Table 1) are consistent with the schematic
cross-section at the International border shown in Figure 3-6 and the geological
cross-sections shown on Figures 3-8 (SERNAGEOMIN (Chile), 2019). This is
significantly different from the DHI, 2018 cross-section drawn to represent
schematically the geology along the same section across the Silala River ravine at
the international border. The DHI interpretation is not supported by the Chilean
data. This means that the layering incorporated in their models, which is used to
represent the hydrogeological configuration, is not supported by the Chilean
evidence. DHI have used the degree of welding in the ignimbrites and their
proposed fault system (see below) to determine their layering and parameter
distributions in their models, which include areas at the international border. The
likely effect of this is that the distribution of hydraulic parameters, both with
depth and laterally, will be wrong, thus affecting the groundwater flow regime
downstream and possibly upstream of the Bolivian wetlands. This would be
highly likely to affect spring flows and the driving groundwater heads in their
models.
The second is that there is no evidence for a major vertical fault running down the
line of the Silala ravine into Chile. DHI assume a vertical fault system, which
provides high hydraulic conductivities to great depth, and has an apparently
arbitrary width, approximately the same as the width of the Silala River ravine at
the international border. This is a major control on groundwater flows in their
model but, as noted, is unsupported by evidence.
The third is that although we recognize that the ignimbrite deposits, where seen at
outcrop in the Bolivian wetlands or in cores, may be highly fractured, and this
would be likely to provide high fracture permeability, the DHI modelling takes no
account of the vertical variability of permeability, as demonstrated by the artesian
201
35
flowing conditions at Chilean borehole SPW-DQN, which implies a significant
confining layer.
Finally, DHI, in their models, do not consider the downstream Chilean geology,
including the controlling influence of the faulting or the presence of low
permeability Pliocene dacitic lavas overlying the Cabana Ignimbrite and
underlying the Chilean Silala Ignimbrite.
The numerical model constructed by DHI was used to make predictions of surface
water and groundwater flows so that they might understand what these flows
might have been before construction of the channels in the Bolivian wetlands. The
flaws in the representation of the hydrogeology by this model are clear. It does not
represent the geology correctly either stratigraphically or structurally and invokes
a fault system that is both unmapped and geometrically highly unlikely. Thus,
DHI assume in their model a distribution of high hydraulic conductivity in the
region of this assumed fault system that has no basis.
The effects of these flaws on the performance of a groundwater numerical model
are unknown but they would undoubtedly mean that actual groundwater flow
paths and the distribution of high and low hydraulic conductivity would be
significantly different to those modelled by DHI. Further evidence of an incorrect
understanding of the hydrogeology is provided by hydrogeochemistry, which is
discussed below in Section 4.
4 HYDROGEOCHEMISTRY OF THE SURFACE AND
GROUNDWATERS OF THE SILALA BASIN
In this section we discuss the importance of the interpretation of the
hydrogeochemistry data for understanding the origins of the groundwater feeding
the various springs in the Silala groundwater catchment and particularly
highlight the differences between the chemistry and isotopic compositions of
202
36
Bolivia’s Cajones and Orientales spring waters. These indicate different origins,
which, though accepted by DHI in their report (BCM, Vol. 4, p. 94), have not
been represented in their modelling.
4.1 Introduction
The study of the chemical and isotopic evolution of surface and groundwaters can
contribute to the understanding of the complex interactions between the river and
the groundwater and the mechanisms of local and regional recharge to the river
flow, differentiating the origins of waters and the residence times that
groundwaters may have spent within the aquifers.
For instance, the chemistry of groundwaters depends on the flow path and the
chemistry of the rocks the water is flowing through. Groundwaters are usually
more mineralized than waters that have run off down a hillslope, reaching rivers
after flowing short distances overland within a short time frame. In the case of the
Silala catchment, the dominant origin for the surface water found in the Silala
River is from groundwater emerging from springs. Groundwaters may spend
many years flowing very slowly through an aquifer. If the spring water has its
origins in one particular aquifer or another, this will often be reflected in the
chemistry of the water. Additionally, groundwater recharge may have isotopic
signatures that reflect the elevation of the precipitation that generated the
recharge.
In this context, the hydrogeochemical study of groundwater has been an important
approach to understand the flow of groundwater and to validate or discard
hypotheses about the conceptual hydrogeological model. In the case of the Silala
basin the study of the hydrogeochemistry has assisted in determining two different
aquifer types and most importantly a differentiation between the spring water of
the Cajones and Orientales wetlands.
203
37
Below we discuss the results of chemical analyses, including isotope studies, of
the Silala River water, spring water and groundwater samples in Chile. These are
presented in detail in Herrera and Aravena, 2019a and 2019b. The samples were
mainly collected over three sampling campaigns during 2016 and 2017 (Herrera
and Aravena, 2019a), to cover both wet and dry seasons. A further smaller
campaign of sampling and analysis of surface and groundwaters was carried out in
2018 in the Quebrada Negra wetland (Herrera and Aravena, 2019b), in which one
spring water, two surface water samples and four groundwater samples from
piezometers were analyzed for the main ion chemistry. There was insufficient
time to carry out isotope analyses on samples collected in this latter campaign.
The results cited by DHI in their report (DHI, 2018) in support of the BCM are
also used to establish the character and origins of the waters of the Silala basin.
The data provided in DHI (BCM, Vol. 4, pp. 89-94) comprise 14 chemical
analyses of water samples from springs and shallow groundwater (sampled from
piezometers) in the Silala River basin in Bolivia. No analyses were reported for
Silala River water. The samples were collected during campaigns carried out for
different studies between the years 2000-2001 and 2016-2017 (BCM, Vol. 4,
pp. 539-542).
Herrera and Aravena (2019a), only considered analyses that had less than 10%
ionic balance error (Custodio and Llamas, 1983). This is common practice for
quality control (Herrera and Aravena, 2019a) and so only 6 of the 14 Bolivian
analyses could be used for comparison with the Chilean data. These included
samples from the Cajones ravine and the Orientales area.
The spatial variation of the chemical composition of the waters can be visualized
using Stiff diagrams. These consist of a polygonal shape of three parallel
horizontal axes extending on either side of a vertical zero axis. Cations are plotted
in milliequivalents on the left side of the zero axes, one to each horizontal axis,
and anions are plotted on the right side. Stiff diagrams were plotted on Figure 4-1
204
38
for all the chemical analyses of the dry season for Chile (the wet season data are
similar) and all Bolivian analyses with satisfactory ionic balance errors.
Figure 4-1. Modified Stiff diagrams of the waters from Silala River area in Chile (rainy
season) and Bolivia (Herrera and Aravena, 2019b).
4.2 Discussion of chemistry analytical results
All the water analyses presented in Herrera and Aravena (2019a) have a relatively
low salinity, though there are significant salinity differences between different
waters.
Figure 4-1 shows that the waters from springs in the northern part of the Silala
River in the Bolivian territory (Cajones ravine and slopes of Cerro Inacaliri) are
characterized by low salinity, ranging between 113 and 129 μS/cm (Herrera and
Aravena, 2019a), similar to that of the springs located in the northern part of the
c::800 \ 1600 / r Me~r(
{ IAmator Proi·.a too,WGS 84
5-'J ' :SI-J DSi4S
Milita : ' /o• ;
•
") . si-a •os-s
Na
Mg
Ca
SCALE
3 meq ! 3 meq
Sample sites
• Springs
• Wells
~ River
Cl
so,
• Surface runoff
205
39
Silala River in the Chilean territory. The groundwater in the Cajones ravine,
collected from shallow piezometers, also has a low salinity, similar to the spring
waters in Chile (Herrera and Aravena, 2019a). All these samples can be seen in
Figure 4-1 to be Na-Ca bicarbonate type.
In contrast, more saline spring waters, ranging between 254 and 394 μS/cm, are
found in the Orientales wetland in Bolivia (Herrera and Aravena 2019a).
Similarly, the groundwater in the Orientales wetland, collected from shallow
piezometers, has a relatively high salinity, similar to that of the Orientales springs.
It is notable that the Orientales spring waters have much higher salinity than the
springs in Chile or in the Cajones area of Bolivia and their conductivities are in
the same salinity range as the groundwater in Chile (including those groundwaters
from the Quebrada Negra wetland). These waters also tend to be Ca-bicarbonate
water type, as do the groundwaters sampled in the Chilean territory.
In Chile, the springs on the northern side of the Silala ravine show similar
chemistries to the river waters and are distinctly different from the deep
groundwater sampled in Chile or the spring water samples in the Orientales
wetland. Downstream of the confluence of the Quebrada Negra with the Silala
River, the river water chemistry has a significantly higher Magnesium content
(see Figure 4-1 and Herrera and Aravena, 2019b). This reflects the contribution of
Magnesium-rich waters found in groundwater samples from the Quebrada Negra
valley and is significantly different to the other waters, including the Bolivian
samples, indicating the extreme complexity of the hydrogeology and origins of
these waters.
Further inspection of Figure 4-1 shows that the spring waters and groundwaters of
the Cajones and slopes of Inacaliri have very similar Stiff diagram shapes, and
salinity, to those springs on the downstream northern side of the Silala River
ravine.
206
40
4.3 Isotope analyses
This section focuses on the evaluation of environmental isotope data collected
from springs, river and wells in Chile in the Silala River topographic catchment.
The stable isotopes referred to in this section are 18O (Oxygen-18),
2H (Deuterium) and 13C (Carbon-13), together with the radioactive isotope of
14C (Carbon-14).
4.3.1. Interpretation of Oxygen-18 and Deuterium (δ18O and δ2H) data
The methodology for the interpretation of the δ18O and δ2H data (Oxygen-18 and
Deuterium) is explained in detail in Herrera and Aravena (2019a). The results are
shown in Figures 4-2 and 4-3, where data for rainy and dry seasons are plotted
with the global meteoric water line (GML) and the local meteoric water line
(LML).
A clear pattern can be seen in these plots. The springs located in the upper course
of the river in Chile (upstream of the junction of the Quebrada Negra with the
Silala River) have a different isotopic fingerprint from the springs located in the
northern part of the lower course of the river (downstream of the junction of the
Quebrada Negra with the Silala River). The data from the latter plot near the local
meteoric water line, which indicates local recharge, whereas the data from the
former plot below the local meteoric water line, indicating recharge from higher
elevations. The results also show that some springs located in the southern part of
the lower river course (downstream of the junction with the Quebrada Negra) in
Chile have a similar isotopic fingerprint to those from the upper river course. This
pattern suggests that these springs are part of the same (or similar)
hydrogeological system as that which feeds the springs in the upper course of the
river in Chile. This is important when the chemistry and isotope data are
integrated for Chile and Bolivia, in terms of the origins of the waters from the
Bolivian springs.
207
41
Figure 4-2. Plot of δ18O and δ2H for river, spring water and wells water in
the rainy season (Herrera and Aravena, 2019a).
-13 -12
PW-BO-
8180 (%0) SMOW
-11
,6_R-Rfo 1
R-Sl-2 PW-OQN-B
SP-Sl-1
MWL-DQN-A
W-80-B
W-80-A
PW-DQN-SI Legend
.6. Silala River rainy season
♦ Springs Silala River rainy season • Wells in Silala River rainy season
-10
-70
-72
-74
-76
-78
-80
-82
~
0
-84 :iii
en
-86
0
~
::c
-88 ~
-90
-92
-94
-96
-98
-100
208
42
Figure 4-3. Plot of δ18O and δ2H for river, spring water and wells water in
the dry season (Herrera and Aravena, 2019a).
Spring water from a Quebrada Negra spring, SP-SI-10, has an isotopic
composition in the range of the spring water issuing from the northern side of the
river course upstream of the junction with the Quebrada Negra.
The isotopic data for the groundwater in both seasons also have a similar
fingerprint, which suggests that all these waters are associated with recharge areas
at similar altitudes. However, in the dry season the group of groundwaters tends to
be somewhat separate from the spring waters, so it may be that the regional
aquifer is recharged at higher altitude than the river and springs in Chile.
MWL-U
W-UQN-B PW-BO-A
PW-B0-B
R-51-7
1:,,,.
-11.6 -11 .4
1-28
Legend
1:,,,. Silala River dry season
♦ Springs Silala River dry season
• Wells in Silala River dry season
-11
-80
-82
-84
-86
-88
~
0
:E
-90
en
J.
:I:
-92
""'°
-94
-96
-98
-100
209
43
It is clear from these isotope data and the chemical analyses (Herrera and
Aravena, 2019a and 2019b) that both river water and springs in Chile upstream of
the Quebrada Negra are likely to be closely related. Electrical tomography
(Arcadis, 2017) has shown the likelihood of perched aquifers in the Alluvial
deposits (Herrera and Aravera, 2019a; Arcadis, 2017) on the southern side of
Cerro Inacaliri. It seems likely that recharge to these perched aquifers in the
Alluvial deposits, which overly the Chilean Silala Ignimbrite, and possibly the
Cabana Ignimbrite, supplies shallow groundwater to the Silala River springs in
Chile upstream of the Quebrada Negra. The alluvial deposits in which these
perched aquifers are found are undoubtedly contiguous with similar deposits in
Bolivia that would similarly be expected to support perched aquifers. Similarly,
perched aquifer(s) in the widespread andesitic lava flows (SERNAGEOMIN
(Chile), 2017) that outcrop in Bolivia to the north west of Cajones wetland and to
the north of the Orientales wetland are likely to supply the groundwater feeding
the springs in the Cajones and in part the Orientales wetlands in Bolivia.
4.3.2. Carbon-14 and Carbon-13 data
There is detailed discussion in Herrera and Aravena (2019a), of the basis for
Carbon-14 dating and the use of Carbon-13 in correcting for several complicating
features, namely carbon input to groundwater from the soil zone, dissolution of
carbonate minerals and from carbon dioxide from volcanic rocks. Because of the
complications and uncertainty attached to such corrections, Herrera and Aravena
restricted their interpretation of the 14C content of the groundwaters sampled in
the Silala River groundwater catchment area to the use of the percent modern
carbon (pMC) as a tracer to evaluate the river-groundwater interactions and riversprings
interactions. In general, the higher the pMC value the younger the water
will tend to be. The Bolivian dates (BCM, Vol. 4, Table 14, p. 92) are not
believed to be credible, because of these complications.
210
44
The Chilean sampling sites and pMC data for the dry season are shown on
Figure 4-4 together with the Bolivian sites and data. The Bolivian data were taken
from DHI (BCM, Vol. 4, Table 14, p. 92).
Figure 4-4. Distribution of 14C sampling points in the Silala River basin in Chile (in the
dry season) and Bolivia together with values of percent modern carbon (14C pMC)
(Herrera and Aravena, 2019a).
From a pMC value of 31.25 (Figure 4-4) at the international border, 14C increases
as the water flows downstream in the Silala River in Chile. This is attributed to
lateral groundwater contributions from the springs flowing into the river from the
north side of the ravine. Further down-gradient beyond the junction with the
Quebrada Negra, the 14C content of the river decreased to 18.1 pMC. This is
caused by a contribution from groundwater discharge from the artesian well,
sample SPW-DQN-SI-O17, which has a 14C content of 8.36 pMC, i.e. much older
Cajone
Hilo
SIN-lXXIII
FCABlnta
c.o.Hri
t' .. SP.Sl-8-017
Jl 167.44)
~ SP-Sl·S-017 J SP-Sl-32·017 (70.66) PW·UQN-B-Ol7
0 123.96) • (9.96)
COOELCO SPW-OQN-SH)17 \
lntak• ;::---- (8.36) ,,_...£--R-S1-3-017
\ ♦ SP-Sl,}1-017 • \.-_,,,...,,.--- (39.15) :: . .>:\ (29.7Sj SP-5~29-017 ~- r \;r-..__R·S1·8-017 j'I- r R·Sl-7-017 (34.49) '
0
Silo/a ' \ Qt, (39.55)
lnac"'iri 132·53) PW-OQN-A-017 (2 1.82) "
6
,-0'!0 N,,ora
Police,Station PW-OQN-B-017 (22.06) R-Sl-9-0l7 ~SP-S"10-017
(lS. l) (27.57)
Sample sites
• Springs
• Wells
"' River
(14CpMC)
l o
I'
I
800 2400
MIiitary
Post
'·• I Sitt2 •·
125,67)
Cr,,.d.Sffolo
4850m
Slte4
(30,67) •
HitoS/11
Hilo
~'N-LXXV
HttoS/~
211
45
water (Figure 4-4). Downstream of this well, further groundwater contributions
from springs from perched aquifers cause an increase in pMC in the river. There
is no information on 14C from river samples given in DHI, 2018.
In contrast, much higher 14C values were found in samples from the springs to the
north of the ravine downstream of the Quebrada Negra in both the dry and rainy
seasons, ranging between 67.44 and 70.66 pMC. These values are comparable to
the 14C value of 86.29 pMC that was reported for a spring located in Bolivia,
supplying the Cajones wetland. Lower 14C values of 25.67 and 30.67 pMC,
similar to the springs in the Chilean sector, are observed in the springs located in
the Orientales wetland in Bolivia. These springs have higher salinity than the
northern (Cajones) wetland springs, similar to that of groundwaters sampled in
Chile, suggesting that the springs are associated with groundwater discharge of a
regional groundwater flow system. The spring located in the Quebrada Negra
(sample SP-SI-10-O17), which may represent discharge of a regional flow system,
perhaps recharged at higher altitude in Bolivia, has a pMC similar to those springs
in the Orientales wetlands.
The deep groundwaters sampled in Chile have much lower 14C values (Figure 4-4)
than the springs, the Silala River or Bolivian samples. As noted above, the lowest
14C value of 8.36 pMC was obtained from groundwater discharging from the
artesian borehole, sample SPW-DQN-SI-O17. The groundwater flowing from this
borehole would normally be confined beneath the upper layers of the Chilean
Silala Ignimbrite.
DHI (BCM, Vol. 4, p. 103), based on the Bolivian 14C data, suggest an “old age in
the southern wetland (up to ~ 11,000 years) and a significantly younger age in
the northern wetland ( up to ~ 1,000 years)”. While it is certainly correct that
these waters have different origins and chemistries, as shown above, these age
estimates are not correct since they do not take into account the dilution effect due
212
46
to dissolution of carbonates along the groundwater flow system and the potential
input of volcanic CO2 (Herrera and Aravena, 2019a).
4.4 Conclusions concerning the origins of the spring waters in the Silala
groundwater catchment
Clearly, the Silala River water that crosses the International border is closely
related to the spring flows emerging from the Bolivian wetlands. Nevertheless, the
chemical and isotopic analyses reveal that the groundwater flow systems are
complex. The proposal of a perched aquifer or aquifers, which supply the springs
in the area to the north of the Silala River ravine, upstream of the junction with
the Quebrada Negra, is justified considering the difference between the chemistry
of the springs and the deep groundwaters and the fact that the deep groundwater
levels in Chile are well below the elevation of the river. The groundwaters and
surface waters that are found in the Quebrada Negra wetland present further
complexity. They are higher in magnesium than any other others analyzed in the
Silala River basin and seem to influence the chemistry of the Silala River
downstream of the junction with the Quebrada Negra ravine, but their high
salinity would indicate that they may be related to a regional deep aquifer.
The difference in salinity, major ion chemistry and 14C pMC values between the
Cajones spring waters and the Orientales spring waters are marked and indicate
different origins for the two sets of springs in Bolivia.
The chemical and isotope analyses of the spring waters in the Cajones area show
strong similarities to those of the springs found on the northern side of the Silala
River ravine, downstream of the junction with the Quebrada Negra. These latter
waters have δ18O and δ2H consistent with a locally recharged origin.
The chemical and isotope analyses of the spring and shallow groundwaters from
the Orientales wetland indicate a different origin, which seems likely to be a
213
47
mixture of shallow, locally recharged groundwater (probably from the alluvial
deposits or Pleistocene lavas of the Inacaliri and Apagado) and groundwater from
a regional aquifer. Some of the samples showed great similarities to the deeper
groundwaters analyzed in Chile.
It is evident that the chemical character of the spring waters of Orientales and of
Cajones is distinctly different. However, in the DHI Near Field model no account
appears to have been taken of these differences and of the differing origins for the
two sets of spring discharges, so the performance of the model in representing the
spring discharges is likely to be flawed. Hence simulated scenarios to predict what
these spring flows might have been in a natural condition, without channelization
or with a restored wetland, are also likely to be flawed. The recharge to one set of
springs finds its way via groundwater flow paths that are distinctly different from
the other, and hence it is highly likely that the residence times for groundwaters
discharging from these springs would be quite different. Therefore, they cannot
sensibly be modelled as if they are the same and have the same recharge areas and
the same origins. This leads to the conclusion that the modelling is based on an
incorrect conceptual understanding of the groundwater flow regime and will
therefore produce flawed results and predictions.
5 SUMMARY OF THE HYDROGEOLOGY OF THE SILALA
GROUNDWATER CATCHMENT - AREAS OF AGREEMENT AND
DISAGREEMENT BETWEEN CHILE AND BOLIVIA
In this section we summarize the hydrogeology of the Silala groundwater
catchment and hence indicate the deficiencies in DHI’s modelling, which fails to
correctly represent the hydrogeology. This matters because unless the
hydrogeology is represented properly the results of various scenario predictions
are likely to be incorrect.
214
48
In briefly describing the hydrogeology of the Silala basin we bring together the
evidence from the geological mapping, radiometric dating and drilling with the
interpretations of hydrogeochemical analyses in both Chile and Bolivia, which
provide convincing support to Chile’s hydrogeological understanding of the
groundwater flow regime in the Silala basin.
It is clear that there are at least two aquifer types that are active in the catchment
in Chile:
a. A perched aquifer system that is present in the alluvial deposits
overlying the bedrock volcanic formations found in the Silala basin
(Arcadis, 2017).
b. A regional aquifer system formed by a succession of ignimbrite
deposits that is interbedded with fluvial debris flow deposits in Bolivia
and Chile and is recharged from the Silala River groundwater
catchment (Arcadis, 2017; DHI, 2018).
It is also clear that recharge to the groundwater catchment, most of which lies in
Bolivia, that enters the ignimbrite regional aquifer either emerges at the Bolivia
wetland springs or flows within the ignimbrites (in Chile, the Chilean-named
Silala Ignimbrite or the Chilean-named Cabana Ignimbrite). There is a clear
groundwater level gradient to Chile from Bolivia in the ignimbrites (Arcadis,
2017), as agreed by the Bolivian experts. The only way that groundwater in the
regional aquifer provided by ignimbrite succession can reach Chile is either as
surface flow from springs in Bolivia’s Near Field area, in particular the Bolivian
Cajones and Orientales wetland springs, or by flowing as groundwater beneath the
area of the Bolivian Near Field model down the hydraulic gradient to Chile. There
is no other possible route for such groundwater flows because of the edifices of
Cerro Inacaliri and Volcán Apagado, whose roots are built upon low permeability
Miocene Volcanic deposits (SERNAGEOMIN (Chile), 2019).
215
49
The hydrogeochemistry analyses have provided strong evidence to support the
existence of these two distinct aquifer systems, which, for the most part, are not
well connected (Arcadis, 2017; Herrera and Aravena, 2017; Herrera and Aravena,
2019a and 2019b). However, DHI do not include these as separate aquifer systems
in their Near Field model, even though they agree that 14C isotope data show
distinct differences between the waters of the Orientales wetlands and the Cajones
wetlands. The configuration of the geology and hydrogeology of the Silala
groundwater catchment, as developed with strong evidence by Chile, is not
incorporated into the DHI Near Field model.
Recharge from precipitation (both rainfall and snowmelt) infiltrates both aquifer
systems, and groundwater flows to a number of spring systems in Chile (some of
which support the Quebrada Negra wetland (Muñoz and Suárez, 2019) and in
Bolivia, where they support the Cajones and Orientales wetlands (Arcadis, 2017;
Muñoz et al., 2017; DHI, 2018).
This recharge provides the flow to the spring and wetland systems. However, the
detail of the geology is highly complex (SERNAGEOMIN (Chile), 2019). This
means that the groundwater flow paths, the distribution of permeability and
origins of recharge to different spring systems are also complex and not precisely
known.
Although there is agreement between the experts on the existence of a regional
aquifer in ignimbrite rocks, the Bolivian interpretation of the three-dimensional
nature of this aquifer system has been shown to be incorrect in several respects:
• It is clear that the ignimbrite aquifer system identified in Chile (the
Chilean Silala and Cabana Ignimbrites), together with an interbedded
fluvial debris flow (section 3, above) has not been recognized by DHI
in their report (DHI, 2018), nor incorporated into their models.
216
50
• The evidence presented shows that the existence of a major fault
system located beneath the Silala River ravine is wholly implausible.
DHI incorporate this fault system as a particular distribution of high
permeability in their Near Field model, which consequently is based on
an incorrect conceptual understanding of the geology and
hydrogeology.
• The DHI modelling takes no account of the vertical variability of
permeability in the regional ignimbrite aquifer, as clearly demonstrated
by the artesian flowing conditions at the Chilean borehole SPW-DQN,
which implies a significant confining layer.
• The impact on the groundwater flow system in the catchment due to
the faulting at the downstream end of the Silala topographic catchment
has not been considered by DHI in their modelling.
• Finally, the difference in origins of the recharge to the Bolivian
wetland spring systems has not been incorporated in DHI’s Near Field
model (section 4.4 above).
It is clear that the hydrogeology of the groundwater catchment is highly complex
and many of the features identified in Chile have not been taken into account by
DHI in their modelling. Given in particular the subtle nature of the changes
associated with the channelization in Bolivia, and lack of recognition of key
features of the hydrogeology, DHI’s scenario predictions must be seen as severely
flawed.
217
51
6 DISCUSSION ON THE ENHANCEMENT OF SPRING FLOWS IN
THE CAJONES AND ORIENTALES WETLANDS BY THE USE OF
EXPLOSIVES
In this section we briefly discuss the assertion made in the BCM that explosives
were used to develop the spring sources in the Bolivian wetlands. We conclude
that enhancement of spring flows by explosives methods as described in the BCM
would not be possible.
Bolivia refers to the use of explosives to enhance spring discharges (‘Many of the
spring discharge points in Bolivia still clearly evidence the use of explosives’;
BCM, Vol. 1, p. 47).
The only evidence to substantiate this claim is a single photograph (BCM, Vol. 4,
p. 101, Figure 44), which includes the bracketed phrase “(precipitates on rock)”.
This is entirely insufficient for DHI to make the statement ‘Based on the rock
blasting in the area of the many of the springs, the current hydraulic gradients may
have been altered from natural conditions’ (BCM, Vol. 4, p. 101). If rock blasting
had been used to excavate the channels at the spring emergences, it is our opinion
that the effects on the hydraulic gradients would be insignificant.
While rock blasting has been used elsewhere to enhance pumped well yields, it is
in our opinion highly unlikely, given the long history of spring discharges (over
centuries and potentially millennia) and associated natural processes of erosion,
that any blasting, had it occurred, would have had a significant impact on spring
discharges. The BCM cites Driscoll, F.G., 1978 (BCM, Vol. 1, p. 47) as evidence
that blasting can enhance water flows by a factor of 6 to 20. The article they refer
to, concerns the development of deep borehole water supplies in poorly fractured
granites, quartzites and slates, not springs. These rocks are metamorphic and have
undergone considerable changes due to high pressure and temperature. They are
normally very poorly permeable. The ignimbrites of the Silala wetlands, by
218
52
Bolivia’s own evidence, are highly fractured, and have major and minor jointing
(SERGEOMIN (Bolivia), 2017).
The deep (well over 100 metres) boreholes undergoing the cited blasting
development were plugged with sand to direct the blast horizontally. Significant
development of spring flow would not be possible using these methods. Bolivia’s
assertion that explosives have been used to enhance spring discharges is therefore
not credible.
7 CONCLUSIONS
(i) What new evidence has been produced, since Chile submitted its Memorial
in July 2017, concerning the understanding of the geology and hydrogeology of
the Silala River?
New investigations in the Silala River topographic catchment have included field
observation, re-logging of borehole drill cuttings, geological mapping and
radiometric dating of the Chilean-named Silala Ignimbrite and Pliocene lavas.
This new information has revealed a more detailed understanding of the
stratigraphy in Chile and the extensive presence of a debris flow that lies at the
base of the Chile-named Silala Ignimbrite and the upper boundary of the Chileannamed
Cabana Ignimbrite. It has also revealed a major fault in Chile, a few
hundred metres below the junction of the Silala River and the Quebrada Negra
tributary valley. The stratigraphy and this structure have not been considered in
the Bolivian hydrogeological conceptual understanding or incorporated into their
numerical models. DHI introduce a new fault system running through the Bolivian
wetlands and down the Silala River ravine into Chile (DHI, 2018), but no
evidence to support this has been found in Chile.
New hydrogeochemical investigations have revealed the distinct character of the
spring and groundwater of the Quebrada Negra. And in conjunction with the
219
53
Chilean data, Bolivian chemical and isotopic analyses have revealed: a) the
distinctly different recharge origins of the spring water of the Bolivian wetlands,
Cajones (referred to in DHI, 2018 as the North Wetland or Bofedal) and
Orientales (referred to in DHI, 2018 as the South Wetland or Bofedal), and b) the
close similarities of the Chilean spring waters, recharged from perched aquifers,
to the spring and groundwaters of the Cajones wetland in Bolivia. As with the
geological structure and stratigraphy, above, this important difference in recharge
to the two Bolivian springs is not incorporated in DHI’s modelling.
(ii) Does the hydrogeological conceptual understanding and parameterisation
of the numerical models of Bolivia’s expert, the Danish Hydraulic Institute (DHI),
provide an adequate basis to quantify the effects of channelization on the surface
water and groundwater flows from Bolivia to Chile?
The DHI numerical models incorporate an incorrect stratigraphy and an
implausible fault system and take no account of the down-gradient Chilean
geological structure or the difference in origin of the Cajones and Orientales
springs waters. In particular,
a. The ignimbrite aquifer system identified in Chile (the Chilean Silala
and Cabana Ignimbrites), together with an interbedded fluvial debris
flow has not been recognized by DHI in their report (DHI, 2018), nor
incorporated into their models, neither has the vertical heterogeneity in
permeability. This will undoubtedly mean that the groundwater
flowpaths that they simulate as a result of their models’ permeability
distribution will be wrong.
b. The fault system that they propose will also affect the groundwater
flowpaths and the ease with which groundwater can move in their
invoked fault system region.
220
54
c. The different origins of the Cajones and Orientales spring waters are
due to the two distinct aquifer systems identified by Chile (with
considerable supporting evidence (sections 3, 4 and 5), but these have
not been included in the DHI models.
d. The faulting mapped at outcrop downgradient in Chile and the
presence of Pliocene lavas (sections 3 and 5) between the two
(Chilean) ignimbrites (in Chile) which cause a decline in the
permeability of the Cabana and Silala Ignimbrites in Chile has
similarly not been considered.
We conclude that the DHI models do not simulate the groundwater system
properly and are unfit to quantify the effects of channelization in the Bolivian
wetlands or accurately represent the current hydrological system.
(iii) Could the flow from groundwater fed springs in the Cajones and
Orientales springs have been significantly enhanced by the use of explosives?
The evidence for showing that the groundwater-fed springs of the Cajones and
Orientales wetland has been enhanced by explosives is flimsy and the reference to
development of deep borehole yields by explosive methods is inapplicable. The
springs could not have been developed significantly to increase yields by the
explosive methods suggested by Bolivia.
In summary, we have shown that the numerical modelling results that have been
presented by Bolivia to demonstrate the alleged effects of channelization in the
Bolivian wetlands are incorrect. Their models are based on a misrepresentation of
the current hydrological system and the proposed scenarios. In short, with this
conceptual basis, their models could only produce implausible predictions.
221
55
8 REFERENCES
Arcadis, 2017. Detailed hydrogeological study of the Silala River. (Chile’s
Memorial, Vol. 4, Annex II).
Almendras, A.O., Balderrama, Z.B., Menacho, L.M., and Quezada, C.G., 2002.
Mapa geológico hoja Volcán Ollagüe, escala 1:250.000. Mapas Temáticos de
Recursos Minerales de Bolivia. SERGEOMIN, Bolivia.
Blanco, N. and Polanco, E., 2018. Geology of the Silala River Basin, Northern
Chile. Servicio Nacional de Geología y Minería (SERNAGEOMIN). (Chile’s
Reply, Vol. 3, Annex XIV, Appendix C).
Custodio, E. and Llamas, M.R., 1983. Hidrología subterránea. Omega.
Barcelona.
Danish Hydraulic Institute (DHI), 2018. Study of the Flows in the Silala Wetlands
and Springs System. (Bolivia’s Counter-Memorial, Annex 17).
Driscoll, F.G., 1978. Blasting – It turns dry holes into wet ones. Johnson Driller’s
Journal, 3.
Herrera, C. and Aravena, R., 2019a. Chemical and isotopic characterization of
surface water and groundwater of the Silala River transboundary basin, Second
Region, Chile. (Chile’s Reply, Vol. 3, Annex XI).
Herrera, C. and Aravena, R., 2019b. Chemical characterization of surface water
and groundwater of the Quebrada Negra, Second Region, Chile. (Chile’s Reply,
Vol. 3, Annex XII).
Herrera, C. and Aravena, R., 2017. Chemical and Isotopic Characterization of
Surface Water and Groundwater of the Silala River Transboundary Basin, Second
Region, Chile. (Chile’s Memorial, Vol. 4, Annex III).
222
56
Latorre, C. and Frugone, M., 2017. Holocene Sedimentary History of the Río
Silala (Antofagasta Region, Chile). (Chile’s Memorial, Vol. 5, Annex IV).
Muñoz, J.F. and Suárez, F., 2019. Quebrada Negra Wetland Study. (Chile´s
Reply, Vol. 3, Annex XIII).
Muñoz, J.F., Suárez, F., Fernandez. B., and Maas, T., 2017. Hydrology of the
Silala River Basin. (Chile’s Memorial, Vol. 5, Annex VII).
Peach, D.W. and Wheater, H.S., 2017. The Evolution of the Silala River,
Catchment and Ravine. (Chile’s Memorial, Vol. 1).
Ríos, H., Baldellón, E., Mobarec, R. and Aparicio, H., 1997. Mapa Geológico
Hojas Volcán Inacaliri y Cerro Zapaleri, escala 1:250.000. Mapas Temáticos de
Recursos Minerales de Bolivia, SGM Serie II-MTB-15B. SERGEOMIN.
Sellés, D. and Gardeweg, M., 2017. Geología del área Ascotán-Cerro Inacaliri.
SERNAGEOMIN, Carta Geológica de Chile, Serie Geología Básica, Santiago.
(Chile’s Memorial, Vol. 6, Appendix G).
SERGEOMIN (Bolivia), 2003. Estudio de cuencas hidrográficas, Cuenca
manantiales del Silala, Cuenca 20. [Study on Hydrographic Basins, Silala Springs
Basins, Basin 20]. (Chile’s Reply, Vol. 2, Annex 94).
SERGEOMIN (Bolivia), 2017. Proyecto Mapeo Geológico-Estructural del área
circundante al manantial del Silala, Departamento de Potosí. Convenio
Interinstitucional, Servicio Geológico Minero (SERGEOMIN)-DIREMAR. La Paz,
Bolivia. (Chile’s Reply, Vol. 3, Annex XIV, Appendix D).
SERNAGEOMIN (Chile), 2017. Geology of the Silala River Basin. (Chile’s
Memorial, Vol. 5, Annex VIII).
SERNAGEOMIN (Chile), 2019. Geology of the Silala River Basin: An Updated
Interpretation. (Chile’s Reply, Vol. 3, Annex XIV).
223
57
Suárez, F., Sandoval, V. and Sarabia, A., 2017. River-Aquifer Interactions Using
Heat as a Tracer in the Transboundary Basin of the Silala River. (Chile’s
Memorial, Vol. 5, Annex X).
Wheater, H.S. and Peach, D.W., 2017. The Silala River Today – Functioning of
the Fluvial System. (Chile’s Memorial, Vol. 1).
Wheater, H.S. and Peach, D.W., 2019. Impacts of Channelization of the Silala
River in Bolivia on the Hydrology of the Silala River Basin. (Chile’s Reply,
Vol. 1).
Wilson, C. and Houghton, B., 2000. Pyroclast Transport and Deposition. In:
Sigurdsson, B.H., Houghton, B., McNutt, S., Rymer, H. and Stix, J. (eds.),
Encyclopedia of Volcanoes, San Diego CA, Academic Press, pp. 545-554.

225
Statement of Independence and Truth
1. The opinions I have expressed in my Reports represent my true and independent
professional opinion. Where I have relied on the observational and monitoring studies
under my supervision by the Chilean scientific experts, or data supplied to me by the
Republic of Chile, I have noted that in my Report.
2. I understand that my overriding duty is to the Court, both in preparing the two
Expert Reports that accompany the Reply of the Republic of Chile and in giving oral
evidence, if required to give such evidence. I have complied and will continue to comply
with that duty.
3. I have done my best, in preparing the Reports, to be accurate and complete in
answering the questions posed by the Republic of Chile in the terms of reference which are
reproduced in the Reports. I consider that all the matters on which I have expressed an
opinion are within my field of expertise.
4. In preparing these Reports, I am not aware of any conflict of interest actual or
potential which might impact upon my ability to provide an independent expert opinion.
5. I confirm that I have not entered into any arrangement where the amount or
payment of my fees is in any way dependent on the outcome of this proceeding.
6. In respect of facts referred to which are not within my personal knowledge, I have
indicated the source of such information.
7. I have not, without forming an independent view, included anything which has
been suggested to me by others, including the technical team and those instructing me.
Dr. HowardWheater
Hydrological Engineer
24 January 2019
226
Statement of Independence and Truth
1. The opinions I have expressed in my Reports represent my true and independent
professional opinion. Where I have relied on the observational and monitoring studies
under my supervision by the Chilean scientific experts, or data supplied to me by the
Republic of Chile, I have noted that in my Report.
2. I understand that my overriding duty is to the Court, both in preparing the two
Expert Reports that accompany the Reply of the Republic of Chile and in giving oral
evidence, if required to give such evidence. I have complied and will continue to comply
with that duty.
3. I have done my best, in preparing the Reports, to be accurate and complete in
answering the questions posed by the Republic of Chile in the terms of reference which are
reproduced in the Reports. I consider that all the matters on which I have expressed an
opinion are within my field of expertise.
4. In preparing these Reports, I am not aware of any conflict of interest actual or
potential which might impact upon my ability to provide an independent expert opinion.
5. I confirm that I have not entered into any arrangement where the amount or
payment of my fees is in any way dependent on the outcome of this proceeding.
6. In respect of facts referred to which are not within my personal knowledge, I have
indicated the source of such information.
7. I have not, without forming an · endent view, included anything which has
been suggested to me by others, in ing the echnical team and those instructing me.
Dr. Denis Peach
Hydrogeologist
24 January 2019
227
LIST OF ANNEXES TO THE REPLY
OF THE REPUBLIC OF CHILE
VOLUME 2
ANNEXES 92 - 99
ANNEX Nº TITLE PAGE Nº
Annex 92 Letter from the General Manager of FCAB in Chile
to the Secretary of the Board of Directors of FCAB,
7 April 1916
(Original in English)
1
Annex 93 Letter from the General Manager of FCAB in Chile
to the Secretary of the Board of Directors of FCAB,
8 September 1916
(Original in English)
7
Annex 94 Bolivian Geology and Mining Survey
(SERGEOMIN), Study on Hydrographic Basins,
Silala Springs Basin, Basin 20, June 2003
(Original in Spanish, English translation)
13
Annex 95 National Report on the Implementation of the Ramsar
Convention on Wetlands Submitted by the
Plurinational State of Bolivia to the 12th Meeting of
the Conference of the Contracting Parties,
2 January 2015
(Original in Spanish, English translation)
85
228
ANNEX Nº TITLE PAGE Nº
Annex 96 Ministry of the Environment and Water of Bolivia,
Characterization of Water Resources in the
Southwest of the Department of Potosí – Municipality
of San Pablo de Lipez “Wetlands of Silala Valley and
Adjacent Sectors” (Volume II), July 2016
(Original in Spanish, English translation)
113
Annex 97 Note N° VRE-Cs-58/2016 from the Ministry of
Foreign Affairs of Bolivia to the Senior Advisor for
the Americas of the Ramsar Convention Secretariat,
27 July 2016
(Original in Spanish, English translation)
187
Annex 98 Ana Paola Castel, Multi-Temporal Analysis through
Satellite Images of the High Andean Wetlands
(bofedales) of the Silala Springs, Potosí – Bolivia,
September 2017
(Original in Spanish, English translation)
193
Annex 99 99.1 Note from the Agent of the Republic of Chile to the
Agent of the Plurinational State of Bolivia,
5 November 2018
(Original in English)
293
99.2 Note from the Agent of the Plurinational State of
Bolivia to the Agent of the Republic of Chile,
22 November 2018
(Original in English)
296
99.3 Note from the Agent of the Republic of Chile to the
Agent of the Plurinational State of Bolivia,
30 November 2018
(Original in English)
297
229
ANNEX Nº TITLE PAGE Nº
99.4 Note from the Agent of the Plurinational State of
Bolivia to the Agent of the Republic of Chile,
11 December 2018
(Original in English)
299
99.5 Note from the Agent of the Republic of Chile to the
Agent of the Plurinational State of Bolivia,
21 December 2018
(Original in English)
302
99.6 Note from the Agent of the Plurinational State of
Bolivia to the Agent of the Republic of Chile,
11 January 2019
(Original in English)
306
99.7 Note from the Agent of the Plurinational State of
Bolivia to the Agent of the Republic of Chile,
7 February 2019
(Original in English)
307

231
LIST OF ANNEXES TO THE
EXPERT REPORTS
VOLUME 3
ANNEXES XI - XIV
ANNEX Nº TITLE PAGE Nº
Annex XI Herrera, C. and Aravena, R., 2019. Chemical and
Isotopic Characterization of Surface Water and
Groundwater of the Silala River Transboundary
Basin, Second Region, Chile
1
Annex XII Herrera, C. and Aravena, R., 2019. Chemical
Characterization of Surface Water and Groundwater
of the Quebrada Negra, Second Region, Chile
69
Annex XIII Muñoz, J.F. and Suárez, F., 2019. Quebrada Negra
Wetland Study
83
Annex XIV SERNAGEOMIN (National Geology and Mining
Service), 2019. Geology of the Silala River: An
Updated Interpretation
187
Data CD CD-ROM containing supporting data to Annexes
XI – XIV
273
Appendix C to
Annex XIV
Blanco, N. and Polanco, E., 2018. Geology of the
Silala River Basin, Northern Chile
273

233
CERTIFICATION
I certify that the annexes and reports filed with this Reply are true copies of
the documents referred to and that the translations provided are accurate.
Ximena Fuentes T.
Agent of the Republic of Chile
15 February 2019