Additional pleading of Chile

INTERNATIONAL COURT OF JUSTICE
DISPUTE OVER THE STATUS AND USE OF THE
WATERS OF THE SILALA
(CHILE v. BOLIVIA)
ADDITIONAL PLEADING OF THE
REPUBLIC OF CHILE
ADDITIONAL PLEADING AND EXPERT REPORT
VOLUME 1 OF 2
16 SEPTEMBER 2019

1
ADDITIONAL PLEADING OF THE REPUBLIC OF CHILE
VOLUME 1
TITLE PAGE Nº
Additional Pleading of the Republic of Chile on Bolivia’s Counter-
Claims
3
Expert Report: Wheater, H.S. and Peach, D.W., Impacts of
Channelization of the Silala River System in Bolivia on the
Hydrology of the Silala River Basin – an Updated Analysis
67
Statements of Independence and Truth of Drs. Howard Wheater
and Denis Peach
149
List of Annexes Volume 2 151
Certification 153

INTERNATIONAL COURT OF JUSTICE
DISPUTE OVER THE STATUS AND USE OF THE
WATERS OF THE SILALA
(CHILE v. BOLIVIA)
ADDITIONAL PLEADING OF THE
REPUBLIC OF CHILE
ADDITIONAL PLEADING AND EXPERT REPORT
VOLUME 1 OF 2
16 SEPTEMBER 2019
3

iii
ADDITIONAL PLEADING OF THE REPUBLIC OF CHILE
LIST OF FIGURES............................................................................................... vi
LIST OF TABLES ................................................................................................ vi
Chapter 1 INTRODUCTION................................................................................. 1
A. Origins and current state of the dispute before the Court............... 1
1. The dispute has been significantly reduced
following Bolivia’s Counter-Memorial.................. 2
2. Bolivia’s position is further weakened in the
Rejoinder ................................................................ 5
B. The structure of the Additional Pleading...................................... 10
Chapter 2 BOLIVIA’S CLAIM TO SOVEREIGN RIGHTS OVER THE
“ARTIFICIALLY ENHANCED FLOW” OF THE SILALA
RIVER HAS NO BASIS IN INTERNATIONAL LAW......................... 11
A. Bolivia’s current claim, that sovereignty over hydraulic
works in its territory creates sovereign rights over the
alleged “artificially enhanced flow” of the Silala River, is
unfounded ..................................................................................... 12
1. Territorial sovereignty does not create
ownership over a shared natural resource ............ 12
2. The legal authorities invoked by Bolivia in
this context are either irrelevant or do not
support its case ..................................................... 16
3. Bolivia’s claim to sovereign rights over the
“artificially enhanced flow” of the Silala
River runs counter to the principles reflected
in the Convention on the Law of the Non-
Navigational Uses of International
Watercourses, in particular, the principle of
equitable and reasonable utilization ..................... 19
4. Bolivia’s right to dismantle the channels is
not exempted from its obligation under
customary international law not to cause
significant harm.................................................... 24
5
iv
B. Bolivia continues to ignore key historical facts, most
importantly its own authorization of the channels in
Bolivian territory .......................................................................... 31
1. Bolivia wrongly suggests that the 1928
channels already existed at the time of the
1906 Chilean concession...................................... 32
2. Bolivia cannot create obligations for Chile,
including the obligation to negotiate
“delivery” and compensation of an alleged
“artificial flow”, by its own actions and/or
omissions in relation to the channels in its
territory................................................................. 32
3. Any damage caused to the wetlands in
Bolivian territory is attributable to Bolivia,
not Chile ............................................................... 34
C. Conclusion: Bolivia’s claim to sovereign rights over the
“artificially enhanced flow” of the Silala River is untenable
under international law and Bolivia’s second and third
Counter-Claims must be dismissed .............................................. 38
Chapter 3 FURTHER ANALYSIS OF THE DHI NEAR FIELD
MODELLING RECONFIRMS THAT BOLIVIA’S
CONTENTIONS ON THE ALLEGED IMPACT OF THE
CHANNELIZATION ARE UNTENABLE AS A MATTER OF
FACT........................................................................................................ 39
A. DHI’s response to the fundamental criticism of Chile’s
experts of the Near Field Model set-up is insufficient or
incorrect ........................................................................................ 40
1. DHI acknowledges that the model results of
the three scenarios (Baseline, No Canal and
Undisturbed) are most likely affected by the
Near Field boundary conditions ........................... 40
2. DHI’s characterization of Chile’s experts
analysis as “vastly simplified” is misplaced ........ 43
B. The DHI modelling data submitted on 7 February 2019
reveal further flaws in the Near Field Model which render
the modelling results wholly unreliable........................................ 48
6
v
1. DHI uses different models for the Baseline
Scenario on the one hand, and the No Canal
and Undisturbed Scenarios on the other
hand, which make them not directly
comparable ........................................................... 48
2. The modelling conditions, including surface
topography and inflow, are different in each
scenario, invalidating any comparison of the
modelling results .................................................. 49
3. Bolivia’s reliance on the 1922 Fox flow
estimate to confirm the results obtained from
the DHI modelling is not credible ........................ 52
C. Conclusion: The DHI Near Field modelling is so severely
flawed that its results are meaningless and should be
dismissed by the Court. ................................................................ 54
SUBMISSIONS ................................................................................................... 57
7
vi
LIST OF FIGURES
Figure 1. Ground surface elevations used in the four Bolivian models compared at
two cross sections of the main channel near the international border (Muñoz et al.,
2019, p. 30, Figure 4-3. CAP, Vol. 2, Annex XV)............................................... 50
LIST OF TABLES
Table 1. DHI Sensitivity Analysis, p. 32, Table 7-2. BR, Vol. 5, p. 80. ............. 42
8
1
CHAPTER 1
INTRODUCTION
1.1 This Additional Pleading is submitted in accordance with the timelimits
fixed by the Court in its Order of 18 June 2019, directing the submission of
an Additional Pleading by the Republic of Chile, limited to the counter-claims
presented by the Respondent.
A. Origins and current state of the dispute before the Court
1.2 The dispute between Chile and Bolivia over the Status and Use of the
waters of the Silala originated in 1999, when Bolivia first denied the international
nature of the Silala River.1 Prior to that, and for almost a century, both Bolivia
and Chile had consistently recognized that the Silala River is an international
watercourse.2
1.3 Bolivia’s change of position in 1999 was not based on any new
scientific evidence. Then, between 2000 and 2009, Chile made efforts to engage
Bolivia in joint technical studies of the Silala River, in the reasonable expectation
that this would allow Bolivia to accept the obvious and previously established
legal fact that the River is an international watercourse.3 These attempts
ultimately failed, due to Bolivia’s insistence on exclusive ownership of the waters
1 Chile’s Memorial (henceforth “CM”), para. 3.8. See 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.
2 See for a full review of the conduct of the Parties in recognition of the international nature of the
Silala River, CM, paras. 4.11-4.66.
3 CM, paras. 3.16-3.25.
9
2
of the Silala and its alleged right to compensation for Chile’s century-long use of
the waters (the so-called “historic debt”). 4
1.4 In March 2016, Bolivia’s President Evo Morales announced that Chile
was “stealing” the waters of the Silala and that Bolivia would present a claim
before the Court. This was followed by statements of the Bolivian Minister of
Foreign Affairs that the presentation of such claim would take at least two years.5
Under these circumstances, Chile decided to present a claim before the Court in
June 2016, to obtain legal certainty over the nature and use of the Silala River as
promptly as possible.
1.5 Only after Chile initiated the present proceedings before the Court did
Bolivia for the first time commission scientific studies of the Silala, to be
undertaken by the Danish Hydraulic Institute (DHI), acting as Bolivia’s expert in
the case.6 This merely confirms that Bolivia’s 1999 “volte-face” had no scientific
underpinnings but was politically motivated.
1. The dispute has been significantly reduced following
Bolivia’s Counter-Memorial
1.6 The dispute was significantly reduced following Bolivia’s Counter-
Memorial of 3 September 2018, in which Bolivia acknowledged the international
nature of the Silala River, and that both States have rights and obligations under
customary international law with respect to the equitable and reasonable
utilization of its water, the obligation not to cause significant harm, and other
4 CM, paras. 3.24-3.25.
5 CM, para. 1.8.
6 Bolivia’s Counter-Memorial (henceforth “BCM”), para. 13.
10
3
procedural obligations under customary international law relating to international
watercourses.7
1.7 However, the Counter-Memorial introduced a new contention by
Bolivia, of alleged sovereign rights over a portion of the Silala waters which it
characterized as “artificial flow”.8 This new and wholly untenable contention is
presented as a defence to Chile’s principal case, and purports to exclude part of
the waters of the Silala from the international law regime.9 It is also presented as
Bolivia’s Counter-Claim b), and underpins Counter-Claim c), claiming that any
delivery from Bolivia to Chile of this alleged “artificial flow” is subject to the
conclusion of an agreement between both States.10 Counter-Claim a), by which
Bolivia claims sovereignty over the artificial channels and drainage mechanisms
in its territory,11 is by contrast not contested by Chile and the Court has no
jurisdiction in the absence of a dispute.
1.8 According to Bolivia, the “artificial flow” of the Silala is “produced”
by the man-made channels in Bolivian territory, and as such is not subject to the
principles of customary international law.12 The portion of “artificial flow” over
which Bolivia claims sovereign rights was estimated by Bolivia’s expert DHI in
its report annexed to the Counter-Memorial, at 30-40% of the surface flow that
crosses the international boundary from Bolivia into Chile.13 Chile notes that this
7 BCM, paras. 16 and 18.
8 BCM, para. 120.
9 Ibid.
10 BCM, para. 181 b) (henceforth also “second Counter-Claim”) and c) (henceforth also “third
Counter-Claim”).
11 BCM, para. 181 a) (henceforth also “first Counter-Claim”).
12 BCM, para. 120.
13 BCM, para. 13.
11
4
percentage range was reduced to 11-33% in the DHI sensitivity analysis annexed
to Bolivia’s Rejoinder of 15 May 2019.14
1.9 The key point is that Bolivia’s distinction between “natural” and
“artificial” flow has no basis either in international law or in science and,
moreover, runs counter to the increasing need for optimal use of fresh water
resources.15 As Chile explained in its Reply on the Counter-Claims of 15 February
2019, any augmented natural flow, the so-called “artificial flow”, resulting from
works on Bolivian territory, is solely attributable to acts of Bolivia and could not
be the source of legal obligations for Chile.16 Bolivia cannot somehow benefit
from its own conduct to create new categories of sovereign rights over a portion
of a shared water resource whose use is governed by the principle of equitable and
reasonable utilization and the obligation not to cause significant harm. Chile thus
explained in its Reply on the Counter-Claims that the lack of any legal foundation
for Bolivia’s claims to sovereignty over an “artificial” flow is dispositive of the
case and sufficient to reject Bolivia’s Counter-Claims b) and c) without any
enquiry into the scientific evidence.17
1.10 In addition, however, Chile demonstrated important flaws in the DHI
modelling set-up, resulting in gross exaggeration of the impact of the channels on
the cross-boundary surface flow, by a factor of approximately 20.18 Chile also
explained how a reduction in cross-boundary surface flow as a result of the
14 Danish Hydraulic Institute (henceforth “DHI”), Updating of the mathematical hydrological
model scenarios of the Silala spring waters with: Sensitivity analysis of the model boundaries,
April 2019 (henceforth “DHI Sensitivity Analysis”). Bolivia’s Rejoinder (henceforth “BR”),
Vol. 5, Annex 25.
15 Chile’s Reply (henceforth “CR”), Chapter 2, in particular Section 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”.
16 CR, paras. 2.69-2.71.
17 CR, paras. 1.17 and 3.1.
18 CR, para. 3.34.
12
5
channels in Bolivia (if any) would inevitably be compensated by increased crossboundary
groundwater flow, except for an estimated loss to increased
evapotranspiration of no more than 2%, agreed on by both Parties’ experts.19
1.11 Chile’s engagement with the facts and the expert reports presented by
Bolivia to support its Counter-Claims b) and c) does not show that “there is a need
to look further at the facts” or that the facts are “dispositive of the case”, as
Bolivia now asserts.20 They are not. What is dispositive is whether international
law recognizes the existence of the alleged “artificial flow.” Chile has
demonstrated that it does not.
1.12 Chile has nonetheless submitted additional expert reports, and
requested the data underlying Bolivia’s expert reports, because it takes seriously
its obligation as a Party to this proceeding to set the factual record straight, for the
benefit of the Court and with a view to any future engagements with Bolivia
related to shared water resources.
2. Bolivia’s position is further weakened in the Rejoinder
1.13 In its Rejoinder on the Counter-Claims of 15 May 2019, Bolivia
continues to contend that customary international law on the use of transboundary
watercourses applies “only to the rate and volume of Silala water that flows
naturally across the Bolivian-Chilean border”.21 However, it no longer refers to
the judicial decisions and State practice relied on in the Counter-Memorial that
were conclusively contested by Chile in its Reply, nor does it make any effort to
respond to Chile’s Reply.
19 CR, paras. 3.9 and 3.47. See DHI, Study of the Flows in the Silala Wetlands and Spring System,
2018 (henceforth “DHI Report (2018)”). BCM, Vol. 2, p. 267.
20 BR, para. 7. See also BR, p. 1, heading 1.A, wrongly suggesting Chile’s “change of position” on
this.
21 BR, para. 70.
13
6
1.14 Instead, Bolivia now claims that its sovereignty over the channels in
Bolivian territory “affords” it sovereignty over the “artificial flow” allegedly
generated by those channels.22 This is a non sequitur as the legal regime of
transboundary watercourses does not follow the territorial sovereignty doctrine
invoked by Bolivia. Moreover, the few – and mainly domestic – legal authorities
cited by Bolivia in this context are either irrelevant to the law of transboundary
watercourses, or do not support Bolivia’s case.23
1.15 Bolivia seeks additional support for its claim to sovereign rights over
the so-called “artificial flow” generated by the channels in Bolivia, in the alleged
benefits accruing to Chile from these man-made channels.24 However, Bolivia
seeks to ignore the fact that such benefits, if any, were neither sought nor
requested by Chile. The alleged benefits are solely the result of Bolivia’s
authorization of the 1928 construction of the channels (under the terms of the
1908 Bolivian concession to a British private company), and Bolivia’s subsequent
decision not to dismantle the channels once it unilaterally terminated the 1908
concession in 1997.25 This is a further reason why Bolivia’s position makes no
sense as a matter of customary international law. It follows that Bolivia has no
case and that Counter-Claims b) and c) must be rejected.
1.16 Although Bolivia continues to make its Counter-Claims b) and c),26 it
suggests that these will become moot once Bolivia has dismantled the channels on
its territory, as proposed under Counter-Claim a).27 In addition, Bolivia suggests
22 BR, p. 38, heading 4.B.1.
23 BR, paras. 72-77. The referred authorities will be discussed in section 2.A of Chile’s Additional
Pleading (henceforth “CAP”).
24 BR, paras. 70 and 81-85.
25 Administrative Resolution N° 71/97 by the Prefecture of the Department of Potosí, 14 May
1997. CM, Vol. 3, Annex 46.
26 BR, p. 56 (Submissions).
27 BR, para. 100.
14
7
that this would also make unnecessary any pronouncement by the Court on
Chile’s principal claims concerning the nature and the use of the waters of the
Silala.28 This is incorrect.
1.17 The dismantling of the channels and/or restoration of the wetlands in
its territory depends on actions (or past omissions) by Bolivia. Such actions (or
omissions) cannot somehow render moot the principled disagreement between the
Parties over the legal characterization of all the Silala waters. Indeed, Chile would
not wish to trouble the Court with this Dispute now that Bolivia has recognized
the nature of the Silala River system as an international watercourse, were it not
for Bolivia’s insistence on the legal existence of an alleged “artificial flow” and its
claims to exclusive sovereign rights thereto, contrary to established principles of
customary international law applicable to international watercourses.
1.18 Bolivia also states that it is entitled to dismantle the works in its
territory “in conformity with its own interests and customary international legal
norms governing transboundary watercourses”.29 This is correct and Chile agrees
that Bolivia has the right to decide whether and how far to maintain the channels,
provided it complies with its obligations under international law, namely (i) to
notify, inform and consult Chile with respect to any planned measure and (ii) not
to cause significant harm to Chile as the downstream riparian State.
1.19 Bolivia’s principal concern in relation to Counter-Claim a) seems to
be that Chile’s principled position that it is “entitled to its current use of the
waters of the Silala River”30 means that “Bolivia’s rights to dismantle the artificial
infrastructure could be constrained if its actions resulted in a reduction in the
28 BR, para. 103.
29 BR, para. 84.
30 BR, para. 33, quoting from CM, Submission c).
15
8
current flow regime such that it prevents Chile from enjoying its existing uses.”31
Bolivia misinterprets Chile’s position. Chile says only that the Silala waters it
presently uses, and those uses themselves, are consistent with international law,
i.e., they qualify as being equitable and reasonable vis-à-vis Bolivia. Indeed,
Bolivia has not asserted otherwise.32
1.20 Thus a reduction (if any) of the cross-boundary surface flow resulting
from the dismantling of the channels in Bolivia would not be considered a
violation of customary international law unless the obligations that Bolivia has
accepted were somehow engaged.33 Nor would non-man made contamination of
the Silala water with larva from insects breeding in the wetlands be considered a
violation of these principles.34
1.21 Chile considers that Bolivia’s sovereignty over the channels in Bolivia
is uncontroversial and that there exists no dispute with respect to Counter-Claim
a). It has asked the Court accordingly to declare that it has no jurisdiction over
Counter-Claim a).35 In the alternative, Chile asks the Court to declare that
Counter-Claim a) is moot or must otherwise be rejected, since Bolivia cannot seek
31 BR, para. 34.
32 See BR, para. 23, for Bolivia’s recognition that the waters have been used by Chile thus far and
that Bolivia has the right to use those waters “to the extent that such uses are consistent with the
principle of equitable and reasonable utilization.”
33 BR, para. 20. These obligations include: “‘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.’” Ibid., quoting from CR, para. 1.3.
34 See BR, para. 43, where Bolivia expresses its concern for the possible sanitary effects of
dismantling the channels.
35 As to Bolivia’s argument that Counter-Claim a) is not without object because a decision of the
Court “can affect existing legal rights or obligations of the parties, thus removing uncertainty from
their legal relations” (BR, p. 20, footnote 65, cite to Northern Cameroons (Cameroon v. United
Kingdom), Preliminary Objections, Judgment, 2 December 1963, I.C.J. Reports 1963, pp. 33-34),
Chile maintains that there is no legal uncertainty since both States agree that Bolivia has
sovereignty over its territory and that the referred principles of customary international law apply
to the actions of Bolivia vis-à-vis the channels in its territory.
16
9
from the Court a declaration that is a truism, i.e., that Bolivia has sovereignty over
its own territory.36
1.22 As to the actual impact of the channels in Bolivia on the cross-border
surface flow, the flaws in the DHI model set-up have been further confirmed, and
several important additional shortcomings have been identified, by Chile’s expert
analysis of the modelling data belatedly received on 7 February 201937, and the
additional modelling data related to the DHI sensitivity analysis, requested by
Chile on 27 May 2019 and submitted by Bolivia on 17 June 2019.38 Among the
most relevant finds, Chile’s experts confirm that different topographies were used
for the different scenarios, of up to 7 m difference: “This is an enormous change
in ground level, equivalent to a structure the height of a two-story building
spanning the width of the river valley and extending at least 200 m along the
river.”39 This is far greater than the changes in channel depth and peat growth that
36 As to Bolivia’s contention that a case is not moot when concessions made by one Party do not
dispose of the dispute in its entirety (BR, p. 21, footnote 65, reference to Southern Bluefin Tuna
Case between Australia and Japan and between New Zealand and Japan, Award on Jurisdiction
and Admissibility, Decision of 4 August 2000, UNRIAA, Vol. XXIII, p. 38, para. 46), Chile
maintains that, in the present case, the positions of both Parties as concerns Bolivia’s first Counter-
Claim are such that they dispose of the dispute (if any dispute ever existed) in its entirety: Chile
agrees that Bolivia has sovereignty over the channels in Bolivian territory and Bolivia agrees that
it must comply with customary international law governing transboundary watercourses when it
seeks to dismantle those channels (see BR, paras. 35 and 84).
37 The DHI modelling data were first requested by Chile by letter of 5 November 2018 (CR,
Vol. 2, Annex 99.1), and again by letters of 30 November 2018 (CR, Vol. 2, Annex 99.3) and
21 December 2018 (CR, Vol. 2, Annex 99.5). The requested data were finally submitted by
Bolivia by letter of 7 February 2019. CR, Vol. 2, Annex 99.7. This was too late to be taken into
account by Chile’s experts in the Reply and Chile reserved its right to refer to these at a later time,
see CR, para. 3.4.
38 Note from the Agent of the Republic of Chile to the Agent of the Plurinational State of Bolivia
of 27 May 2019. CAP, Vol. 2, Annex 100.1. Note from the Agent of the Plurinational State of
Bolivia to the Agent of the Republic of Chile of 17 June 2019. CAP, Vol. 2, Annex 100.2.
39 Muñoz et al., Assessment of the Silala River Basin Hydrological Models Developed by DHI,
2019 (henceforth “Muñoz et al. (2019)”), p. 33. CAP, Vol. 2, Annex XV.
17
10
the scenarios were meant to evaluate.40 In the informed opinion of Chile’s experts,
this renders the DHI modelling exercise wholly unreliable and misleading. Indeed,
Chile’s experts are shocked by the basic errors committed by DHI.41
B. The structure of the Additional Pleading
1.23 The structure of this Additional Pleading is as follows: chapter 2
discusses and rejects Bolivia’s claim that its sovereignty over the artificial
channels in Bolivia affords it sovereignty over the “artificial flow” allegedly
generated thereby; in chapter 3 Chile addresses the fundamental flaws in the DHI
model set-up, this time incorporating in its analysis the model data received from
Bolivia on 7 February 2019, as well as the additional model information related to
the DHI sensitivity analysis submitted as Annex 25 of Bolivia’s Rejoinder. This
additional information confirms Chile’s earlier conclusion that the percentage of
so-called “artificially enhanced flow”, if existing at all, is grossly overstated and
the outcome of a fundamentally flawed hydrological model.
1.24 This Additional Pleading is supported by a joint expert report by
Drs. Howard Wheater and Denis Peach, in turn supported by two underlying
studies into the Silala River system that are annexed.
40 Wheater, H.S and Peach, D.W., Impacts of Channelization of the Silala River in Bolivia on the
Hydrology of the Silala River Basin – An Updated Analysis, August 2019 (henceforth “Wheater
and Peach (2019b)”), pp. 8 and 63. CAP, Vol. 1, Expert Report.
41 Wheater and Peach (2019b), p. 40.
18
11
CHAPTER 2
BOLIVIA’S CLAIM TO SOVEREIGN RIGHTS OVER THE
“ARTIFICIALLY ENHANCED FLOW” OF THE SILALA RIVER HAS
NO BASIS IN INTERNATIONAL LAW
2.1 Chile’s Reply devoted a chapter to the proposition that “Bolivia’s
Claims to the ‘Artificially Enhanced Flow’ of the Silala River have No Support in
International Law and Ignore Key Historical Facts.”42 In its Rejoinder, Bolivia
fails to refute, or even contest, Chile’s arguments supporting this proposition.
Nevertheless, Bolivia continues to cling to its invented distinction between
“natural” and “artificial” flows. From all that appears, Bolivia does this because it
insists on adhering to its claim that it has sovereignty, in the sense of ownership,
over something it calls the “artificially enhanced flow” of the Silala, created by
the infrastructure installed by the British company, the Antofagasta (Chili) and
Bolivia Railway Company (FCAB), in the late 1920s pursuant to authorization
Bolivia itself granted.43
2.2 In section 2.A below, Chile responds to Bolivia’s untenable theory
that sovereignty over the channels in Bolivia affords it sovereignty over the waters
allegedly generated by those channels, and the legal authorities relied on by
Bolivia in support thereof. In section 2.B, Chile recalls key historical facts related
to the construction of the channels that Bolivia has chosen thus far to ignore, and
notes the implausibility of Bolivia creating legal rights against Chile based on its
own unilateral acts.
42 CR, Chapter 2, pp. 9-45.
43 See for the history of the 1928 channels, CM, para. 4.61 and CR, paras. 2.61-2.64. See for
Bolivia’s authorization of the construction of the channels, CR, paras. 2.69-2.71. Bolivian
authorization of the channel construction is further addressed in Chapter 2.B below.
19
12
A. Bolivia’s current claim, that sovereignty over hydraulic works in its
territory creates sovereign rights over the alleged “artificially enhanced
flow” of the Silala River, is unfounded
2.3 Before discussing the legal merits, or lack thereof, of Bolivia’s current
claim, it bears emphasis that the artificial “enhancement” (if it exists) of the Silala
surface flow would result not from water brought in by Bolivia (or anyone else)
from outside the natural catchment of the Silala basin (so-called “developed”
water), but from slightly increasing the proportion of groundwater that appears as
surface water in the river channel. As Chile has previously shown,44 and as agreed
by Bolivia’s expert DHI,45 the groundwater would in any event flow within the
aquifer to Chile. Thus, the “enhancement” is fictional: the surface flow may be
slightly increased, but the total quantities reaching Chile would ultimately be
substantially the same.
1. Territorial sovereignty does not create ownership over a shared natural
resource
2.4 In its Rejoinder Bolivia persists in its claim that its “sovereignty over
the artificial infrastructure in its territory affords [it] sovereignty over the artificial
flows generated by that infrastructure.”46 A moment’s reflection demonstrates that
this cannot be the case.47
44 Wheater, H.S and Peach, D.W., Impacts of Channelization of the Silala River in Bolivia on the
Hydrology of the Silala River Basin, January 2019 (henceforth “Wheater and Peach (2019a)”),
pp. 7-8. CR, Vol. 1, pp. 103-104.
45 DHI Report (2018). BCM, Vol. 5, p. 84: “groundwater level gradients and hydrogeological
properties clearly indicate groundwater flow from Bolivia to Chile”.
46 BR, p. 38, heading 4.B.1.
47 Chile addressed this claim by Bolivia in its Reply. CR, paras. 2.26-2.39.
20
13
2.5 None of the water in the Silala River system has been brought in by
Bolivia from outside sources. All of the water is generated naturally within the
Silala groundwater and topographic catchments. The fact that the water runs
through infrastructure in Bolivia, installed with Bolivia’s authorization, cannot
change the fact that it is Silala water. There is nothing “artificial” about it.
2.6 As Bolivia recognizes, Chile has from the beginning accepted
Bolivia’s ownership over the infrastructure in its territory.48 This follows from the
fundamental principle of territorial sovereignty. The infrastructure is fixed in
place and forms part of Bolivia’s territory. Some domestic legal systems would
call it a “fixture” or set of fixtures – things, like buildings, that form part of the
land and pass to the buyer with the sale of land.49 Like Bolivia’s infrastructure,
they cannot be said to be shared with a neighbour.
2.7 But it does not follow that an international watercourse that passes
through the infrastructure, constructed with Bolivia’s authorization and over
which Bolivia has sovereignty, partakes of the sovereign character of that
infrastructure. Sovereignty does not somehow “rub off” of water works onto the
water they carry. This novel theory is not accepted in international law and,
moreover, would be a dangerous and disruptive proposition, given that nearly all
of the world’s major international watercourses have been subjected to some kind
of human intervention.50 More generally, as Chile has demonstrated,51 a State
cannot, by definition, have exclusive sovereignty over something that is shared
48 BR, para. 71.
49 See, e.g., Michael Allan Wolf, 8 Powell on Real Property section 57.04, (Matthew Bender,
17 vols., updated quarterly; ISBN: 9780820515502).
50 Examples abound. To take several regions, there are dams and other works on: the Blue Nile
(Africa), the Mekong (Asia), the Danube (Eastern Europe), the Colorado (North America), the
Paraná (South America) and the Rhine (Western Europe).
51 CR, paras. 2.26-2.39.
21
14
with another State.52 In this case, what is shared is the waters of the Silala River
system. Those waters include those that pass through the system of channels
constructed by the FCAB in the late 1920s. This does not transform them into
something over which Bolivia has sovereignty in terms of ownership.
2.8 As Chile noted in its Reply,53 the River Oder case54 offers useful
insights into the validity of claims of exclusive sovereignty over waters of an
international watercourse that are, for the time being, situated wholly within the
territory of a State. It will be recalled that in that case, despite Poland’s territorial
sovereignty over portions of an international watercourse, the Permanent Court
held that use of the watercourse is not governed by the rule of territorial
sovereignty but by the principle of “a community of interest of riparian States”.55
The Permanent Court explained that “this community of interest […] 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.”56 While the River Oder case involved navigation, the Court applied the
52 Writers have sometimes resorted to the concept of “limited territorial sovereignty” to describe a
State’s rights in shared freshwater resources. See, e.g., F.J. Berber, Rivers in International Law,
pp. 25, et seq., Stevens & Sons, London (1959); L. Caflisch, Règles Générales du Droit des Cours
d’Eau Internationaux, 219 Recueil des Cours (1989-VII) 9, pp. 55, et seq. (1992); and J. Lipper,
Equitable Utilization, in A. Garretson, R. Hayton & C. Olmstead, eds., The Law of International
Drainage Basins, p. 15, at p. 23, et seq., Oceana, Dobbs Ferry, N.Y. (1967). A State’s sovereignty
over its territory is said to be “limited” by the obligation not to use that territory in a way that
harms other States. This principle was recognized in the Trail Smelter Arbitration (United States v.
Canada), 16 April 1938 and 11 March 1941, 13 UNRIAA 1905, at 1965 (1941), which was cited
by the tribunal in the Indus Waters Kishenganga Arbitration (Pakistan v. India), Permanent Court
of Arbitration, Partial Award of 18 February 2013, para. 449.
53 CR, para. 2.30.
54 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), pp. 5-46.
55 Ibid., p. 27.
56 Ibid., p. 27.
22
15
community of interest principle to non-navigational uses in the Gabčíkovo-
Nagymaros case.57
2.9 The “perfect equality” of riparian States in the community of interest
means that each enjoys equality of right vis-à-vis the other in the use of the shared
watercourse. This principle is at the core of the provisions contained in the 1997
UN Convention on the Law of the Non-Navigational Uses of International
Watercourses (UNWC or “the Convention”),58 in particular, the principle that the
allocation of rights of the riparians to use an international watercourse is governed
by the principle of equitable and reasonable use and the obligation not to cause
significant harm.
2.10 The principle of equitable and reasonable use requires that all relevant
factors be taken into consideration, including geographical and hydrological
factors linked to the territorial sovereignty of the riparian States over the portions
of the watercourse located in their territories. Article 6 of the UNWC contains a
non-exhaustive list of relevant criteria, and provides that no criterion has
preference over the others.
2.11 Equality of right and the related principle of equitable and reasonable
use thus preclude Bolivia from asserting territorial sovereignty over Silala waters,
whether “enhanced” by infrastructure or not. Just as Poland enjoyed no greater
57 Gabčíkovo-Nagymaros Project (Hungary/Slovakia), Judgment, I.C.J. Reports 1997, p. 7, at
p. 56, para. 85. The Court said there: “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 on the Law of the Non-Navigational Uses of
International Watercourses by the United Nations General Assembly.” See also Pulp Mills on the
River Uruguay (Argentina v. Uruguay), Provisional Measures, Order of 13 July 2006, I.C.J.
Reports 2006, p. 113, at p. 122, para. 39 and p. 130, para. 64, and Rhine Chloride Arbitration
concerning the Auditing of Accounts (The Netherlands/France), Permanent Court of Arbitration,
Award of 12 May 2014, para. 97.
58 Convention on the Law of the Non-Navigational Uses of International Watercourses (henceforth
“UNWC” or “the Convention”), signed at New York on 21 May 1997, U.N. Doc. A/RES/51/229
(1997). CM, Vol. 2, Annex 5.
23
16
rights in the Oder’s tributaries by virtue of their origin and flow for a significant
distance there, Bolivia has no “preferential privilege” in Silala waters by virtue of
their origin and flow, for a relatively short distance, there.
2.12 The fact that what Bolivia calls the “artificially enhanced flow” of the
Silala is composed of waters of the Silala River, not waters that are somehow
imported, combined with the law governing international watercourses as declared
by the Court, necessarily excludes any assertion of sovereignty, in the sense of a
sole right of control, by Bolivia over any portion of the waters of the Silala.
2. The legal authorities invoked by Bolivia in this context are either
irrelevant or do not support its case
2.13 In its Rejoinder Bolivia attempts to bolster its case for sovereignty
over what it characterizes as the “artificially enhanced” surface flow of the Silala
River by citing various legal authorities. As was the case with the authorities
relied on in Bolivia’s Counter-Memorial, these are of no help to Bolivia because
they are either inapposite to the present case or do not support Bolivia’s argument.
2.14 In its attempt to persuade the Court that a riparian State may claim
ownership over the “artificially enhanced” waters of an international watercourse,
Bolivia relies on two domestic decisions of courts of the State of California,
United States of America (USA), one by an intermediate court of appeal from
201259 and the other by the California Supreme Court from 1908.60
2.15 The first case involved storm water that was stored in a reservoir. The
court classified this stored water as “salvaged” water, which according to
59 City of Santa Maria et al. v. Adam et al., Court of Appeal, Sixth District, California, 211 Cal.
App. 4th 266, 149 Cal. Rptr .3d 491 (2012).
60 Pomona etc. Co. v. San Antonio etc. Co., Supreme Court of California, 152 Cal. 618 (1908).
24
17
California law is “water that is saved from waste as when winter floodwaters are
dammed and held in a reservoir.”61 Under California law salvaged water is owned
by the party that created it. The present case does not involve storm water or any
other form of water saved from waste by being held in a reservoir. It therefore
does not involve “salvaged water” as defined by that California court and law.
This case is thus inapposite to the present one and of no help to Bolivia.
2.16 The second case also involved “salvaged” or “rescued” water. It noted
the principle of California law that as long as a downstream user receives the
quantity of water it is entitled to receive, it may not complain that an upstream
user increased the quantity of water available to it through salvaging or
developing the water.62 This case, also, is of no help to Bolivia. Again, no
“salvaged” or “developed” water is involved here.
2.17 The more fundamental difficulty with Bolivia’s reliance on these
domestic cases is that international law does not recognize the concepts of
“salvaged” or “developed” water. These are domestic law expressions that not
only describe the manner in which the water was collected by a party, but also
carry with them legal rules about the allocation of water rights and the
accommodation of competing uses between water users. These domestic legal
rules do not necessarily follow the rule of equitable and reasonable use that is
applicable to riparian States under international law. Thus the principle stated in
the 1908 California case, that a downstream party may not complain so long as it
receives the pre-determined volumetric allocation to which it is entitled, is
irrelevant to the rights of co-riparians to an international watercourse under
customary international law.
61 BR, para. 74.
62 BR, para. 75.
25
18
2.18 Domestic law cases decided in similar factual situations may possibly
shed light on the implications of claims, such as now made by Bolivia, to except
part of the flow of a river from an otherwise established legal regime. In that
respect, a more apposite domestic law case, in addition to the one discussed in
Chile’s Reply,63 is Southeastern Colorado Water Conservancy District v. Shelton
Farms, Inc.,64 a 1974 en banc decision of the Colorado Supreme Court, USA. In
that case, an upstream riparian landowner, Shelton, cleared areas along the river of
phreatophytes (water-consuming plants) and filled in a marshy area. “Shelton
claimed he had saved approximately 442 acre-feet [545,198 cubic metres] of
water per year, previously consumed by phreatophytes or lost to evaporation,
which is not available for beneficial use. […] He asked for the right to augment
his previous water rights with the salvaged water […].”65
2.19 The Colorado Supreme Court denied Shelton any rights in the water.
It reasoned that “in this situation unrestrained self-help to a previously untapped
water supply would result in a barren wasteland,” because of the incentive such a
right would create to denude river banks of trees and other plant growth.66 But
such self-help could also take the form of installing lined canals or channels to
enhance available water supplies. Like the removal of phreatophytes, this could,
63 This case is 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), discussed in CR, paras. 2.8-2.9. As explained in
the Reply, the case involved facts that are similar to those of the present case.
64 Southeastern Colorado Water Conservancy District v. Shelton Farms, Inc., 187 Colo. 181, 529
P.2d 1321 (1975). This case was referred to in footnote 34 of CR, at p. 13.
65 Ibid., p. 184.
66 Ibid., p. 191. The court memorably stated that if Shelton were to prevail, “the use of a power
saw or a bull-dozer would generate a better water right than the earliest ditch on the river.”
Ibid., p. 191. (The court’s reference to the “earliest ditch” refers to the prior appropriation doctrine
applied in Colorado, in which “first in time is first in right.”)
26
19
in the words of the court, “create a super class of water rights never before in
existence.”67
2.20 Bolivia also refers to the European Union (EU) Water Framework
Directive of 2000.68 It is not clear to Chile how this helps Bolivia, or the Court.
Bolivia quotes definitions of two terms (“Artificial water body” and “Heavily
modified water body”), neither of which appears to apply to the Silala River. The
Silala is not “a body of surface water created by human activity” and is thus not
an “Artificial water body;” nor is it “a body of surface water which as a result of
physical alterations by human activity is substantially changed in character,” and
is thus not a “Heavily modified water body.” Even if the Silala qualified under
one of these definitions, the EU Water Framework Directive, if applicable, quod
non, does not contain rules that would provide guidance in this case. While it is no
doubt an important regional instrument in the field of water management, the EU
Water Framework directive is inapposite to the present case.
3. Bolivia’s claim to sovereign rights over the “artificially enhanced flow” of
the Silala River runs counter to the principles reflected
in the Convention on the Law of the Non-Navigational Uses of
International Watercourses, in particular, the principle of equitable and
reasonable utilization
2.21 The Convention on the Law of the Non-Navigational Uses of
International Watercourses (UNWC) is considered in material parts to reflect rules
of customary international law, and Bolivia has accepted in its pleadings the
67 Ibid., p. 190. A practical difficulty noted by the Colorado Supreme court was the technical one
of determining the amount of water salvaged, or saved, through the actions of upstream users.
This would be an issue in the present case, as well, if Bolivia were to leave the infrastructure
installed by the FCAB in place.
68 BR, para. 76.
27
20
customary nature of the basic obligations reflected in the Convention.69 Foremost
among these obligations is that of equitable and reasonable utilization, recognized
in its Article 5. The Court confirmed the fundamental nature of that principle in its
judgment in the Gabčíkovo-Nagymaros case,70 decided four months after the
conclusion of the UNWC, in which it cited a State’s “basic right to an equitable
and reasonable sharing of the resources of an international watercourse.”71 The
Court in that case also referred to the Convention as evidence of modern
development of international law that has strengthened the principle of
community of interest for non-navigational uses of international watercourses.72
2.22 Assertions of sovereignty over shared freshwater resources, like the
Silala River, in the sense of property rights, are antithetical to the principles
expressed in the UNWC, and indeed, everything the Convention – and thus, the
corresponding rules of customary international law – stands for. It is impossible to
reconcile the concept of a community of interest with assertions of exclusive
sovereignty over that in which such a community exists.
2.23 The only time the word “sovereign” is used in a substantive provision
of the Convention is in Article 8, on the General Obligation to Cooperate. Article
8(1) of the Convention provides that: “Watercourse States shall cooperate on the
basis of sovereign equality, territorial integrity, mutual benefit and good faith in
order to attain optimal utilization and adequate protection of an international
watercourse.”73 Thus, “sovereign” equality is a basis of cooperation, not a basis
for a State to go its own way.
69 This may be inferred from Bolivia’s repeated reliance on the Convention, and its preparatory
work, beginning with its first written pleading. See BCM, chapter 3, pp. 57 et seq.
70 Gabčíkovo-Nagymaros Project (Hungary/Slovakia), Judgment, I.C.J. Reports 1997, p. 7.
71 Ibid., p. 54, para. 78.
72 Ibid., p. 56, para. 85.
73 UNWC, art. 8(1). CM, Vol. 2, Annex 5.
28
21
2.24 The “artificially enhanced flow” over which Bolivia asserts
sovereignty, is solely Silala water. It originates, and forms part of, the Silala River
system. It was not brought in by Bolivia from outside the basin. It does not even
consist of captured storm water, such as that involved in some of the domestic-law
cases on which Bolivia relies. It is Silala River water, even if it passes through
infrastructure that, according to Bolivia, maximizes surface flow. As such, it is
governed by the rules of customary international law reflected in the UNWC.
2.25 Bolivia in its Rejoinder challenges Chile’s reference to Article 25 of
the UNWC, entitled “Regulation.”74 Chile explained in its Reply that regulation of
watercourses in the sense of Article 2575 is a very common practice, and that
many forms of such regulation may have the effect of optimizing surface flows –
whether through lining or straightening channels, fortifying river banks, or other
such works. And yet these works have not heretofore to Chile’s knowledge given
rise to claims of sovereignty over optimized flows.
2.26 Bolivia asserts that “Chile does not engage in any effort to show how
[Article 25] applies to non-state parties to the Convention.”76 Apart from the fact
that the relevant provisions of the UNWC are recognized as a reflection of the
basic rules of international law governing the non-navigational uses of
international watercourses, something Bolivia has accepted in its pleadings,77
Chile’s point in referring to Article 25 was not so much to deploy it as a reflection
of a legal rule as to indicate a general and widespread practice of installing
74 BR, para. 72. Chile refers to Article 25 UNWC in CR, para. 2.7.
75 Article 25 defines the term “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.” UNWC, art. 25(3). CM, Vol. 2, Annex 5.
76 BR, para. 72.
77 See BCM, para. 133, for Bolivia’s recognition that it is bound by those provisions of the UNWC
that reflect customary international law.
29
22
infrastructure in international watercourses that may have the effect of
augmenting the volume of those watercourses.
2.27 Bolivia maintains that Article 25 “was never intended to address the
augmentation of the volume of the watercourse flow through artificial works, but
only […] the augmentation of the efficiency and quality.”78 Yet the Convention’s
definition of the term “regulation” leaves ample room for flow augmentation
through artificial works. 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.”
Variation of the flow could obviously be by way of augmentation.
2.28 Indeed, it is clear from the travaux preparatoires that the International
Law Commission took into account, among other things, the set of nine articles on
the regulation of the flow of international watercourses adopted by the
International Law Association (ILA) in 1980 in preparing Article 25.79 The ILA’s
articles define “regulation” to mean: “continuing measures intended for
controlling, moderating, increasing or otherwise modifying the flow of the waters
in an international watercourse for any purpose; […] ” (emphasis added).80
Nothing in the ILC’s commentary to Article 25 indicates disagreement with this
definition.
78 BR, para. 72.
79 The ILA’s articles are referred to in the ILC’s commentary to the U.N. Convention, 1994
Yearbook of International Law, vol. II, part two, p. 127, footnote 393. This footnote contains a
cross-reference to the Fifth Report of the Special Rapporteur, where the ILA’s articles are set forth
in full. Stephen C. McCaffrey, Fifth Report on the Law of the Non-Navigational Uses of
International Watercourses, 1989 Yearbook of International Law, vol. II, part 1, p. 91, at p. 125
(henceforth “Fifth Report”). It is clear from the Convention’s travaux that the ILC’s work was
generally influenced by that of the ILA, particularly the ILA’s Helsinki Rules on the Uses of the
Waters of International Rivers, ILA, Report of the Fifty-Second Conference, Helsinki, 1966,
p. 484, ILA, London, 1966.
80 Fifth Report, p. 125.
30
23
2.29 There is therefore no evidence that either the ILC or the diplomatic
conference that negotiated the UNWC on the basis of the ILC’s articles81 intended
to (a) exclude increased flows resulting from regulatory works from the concept
of “regulation,” or (b) accept that the State in which regulatory works are located
has sovereignty over any increased flows. This confirms the conclusion that flow
augmentation is not excluded from the ambit of Article 25.
2.30 In addition, Bolivia’s argument that Article 25 was intended to
address “only […] the augmentation of the efficiency and quality” of the volume
of the watercourse flow through artificial works, not “the augmentation of the
volume of the watercourse flow through artificial works,”82 collapses on itself
since augmentation of efficiency would ordinarily have the effect of increasing
volume.
2.31 The use of efficiency as part of an equitable utilization analysis is
anticipated by Article 6 of the UNWC, which in subparagraph (f) refers to
“[c]onservation […] and economy of use of the water resources of the
watercourse” as possibly relevant factors. This would make sense particularly in
arid environments such as that of the Silala River, which is located in one of the
driest areas on Earth. More efficient use leaves more water to be used by the
riparians. But contrary to Bolivia’s assertion, a State implementing more efficient
methods to use water does not thereby acquire ownership over the surplus waters
generated by more efficient methods of use.
2.32 As a consequence of these rules of international law, Bolivia has an
obligation to utilize the waters of the Silala River system in an equitable and
81 At the conclusion of its work, the conference adopted the following “Statement of
Understanding”: “Throughout the elaboration of the draft Convention, reference had been made to
the commentaries to the draft articles prepared by the International Law Commission to clarify the
contents of the articles.” U.N. Doc. A/51/869, 11 April 1997.
82 BR, para. 72.
31
24
reasonable manner. Equitable and reasonable utilization cannot be determined
unilaterally or in the abstract. It must be arrived at through consideration of all
relevant factors by the States involved,83 with a view to achieving a solution that
is fair, balanced, and reasonable.
4. Bolivia’s right to dismantle the channels is not exempted from its
obligation under customary international law not to cause significant
harm
2.33 In a rather confused section of its Rejoinder entitled “Bolivia’s Right
to Dismantle the Artificial Infrastructure and Significant Harm Considerations,”84
Bolivia commits to conduct any work it may undertake to preserve the wetlands in
Bolivia, including the possible removal of the stone-lined channels installed in the
1920s, in a manner consistent with its obligations under international law: “If
Bolivia sought to dismantle the artificial infrastructure that was installed within its
territory and return the Silala to its natural, pre-artificial state, it would do so in
accordance to its rights and obligations under international law and in a manner
that does not create significant transboundary environmental harm.”85
2.34 However, having recognized the obligation not to cause significant
transboundary harm, and having pledged to follow it, Bolivia proceeds to state
that “it is [the obligation’s] application to the particular circumstances of the
Silala that should be further clarified.”86 Bolivia then states that “[t]he obligation
not to cause significant harm must be determined proportionally by balancing
83 UNWC, art. 6. CM, Vol. 2, Annex 5.
84 BR, p. 16, heading 3.B.2.
85 BR, para. 35.
86 Ibid.
32
25
against the rights of the acting State to pursue its own interests and priorities, such
as development and environmental protection and restoration.”87
2.35 What Bolivia intends by this sentence is not entirely clear. Bolivia
seems to be saying that the obligation not to cause transboundary harm must be
balanced “against the rights of the acting State to pursue its own interests and
priorities, such as development and environmental protection and restoration.”88 If
taken literally, such an interpretation could defeat the “no-harm” obligation itself,
as it would allow a State to observe the no-harm obligation unless and to the
extent that its interests and priorities dictate otherwise. This amounts, once again,
to seeking refuge in sovereignty, in effect asserting a sovereign right in Bolivia to
do essentially whatever it believes is in its best interests, notwithstanding rules of
international law that dictate otherwise.
2.36 Bolivia seeks support for its interpretation of the no-harm obligation
first from the arbitral award in the Indus Waters Kishenganga Arbitration and
second from the Court’s judgment in the Pulp Mills on the River Uruguay case.
2.37 Chile would first note that both of these cases involved treaties and the
interpretation thereof, the Indus Waters Treaty of 196089 and the Statute of the
River Uruguay of 1975.90 Both treaties establish joint commissions between the
States involved and are known for being well-crafted and, especially in the case of
the Indus Waters Treaty, highly detailed. General lessons from the cases should
therefore be drawn with caution.
87 BR, para. 36.
88 Ibid.
89 Indus Waters Treaty, Karachi, 19 September 1960, entered into force on 1 April 1960, 419
UNTS, No. 6032, p. 125.
90 Statute of River Uruguay, 26 February 1975, entered into force on 18 September 1976, 1295
UNTS, No. 1-21425, p. 340.
33
26
2.38 With respect to the Kishenganga case, Bolivia deploys the case in an
effort to support its assertion that the no-harm rule should be balanced against the
acting State’s interests and priorities in the field of development, in particular.91
Bolivia states that the Kishenganga Court of Arbitration “asserted that, ‘[t]he
requirement to avoid adverse effects on Pakistan’s agricultural and hydroelectric
uses of the waters of the Kishenganga/Neelum cannot, however, deprive India of
its right to operate the [Kishenganga Hydroelectric Plant]”.92 Interestingly, this
passage appears in a section of the Tribunal’s Partial Award entitled “(b) The
preservation of downstream flows.”93
2.39 Unfortunately, Bolivia offers no explanation as to how the quoted
passage helps its case. The requirement and right referred to in the quoted passage
derive not from customary international law, but from the Indus Waters Treaty,
specifically, Paragraph 15(iii) of Annexure D thereof. In the words of the Court of
Arbitration, this provision “gives rise to India’s right to construct and operate
hydro-electric projects involving inter-tributary transfers but obliges India to
operate those projects in such a way as to avoid adversely affecting Pakistan’s
‘then existing’ agricultural and hydro-electric uses.”94 India is upstream on the
Kishenganga River and Pakistan downstream. Thus, to the extent that the award is
at all applicable to the present case, Bolivia would be in the position of India and
Chile in that of Pakistan, and the quoted passage would favour Chile.
91 BR, para. 36.
92 Ibid.
93 Indus Waters Kishenganga Arbitration (Pakistan v. India), Permanent Court of Arbitration,
Partial Award of 18 February 2013, p. 168.
94 Ibid.
34
27
2.40 Bolivia then turns to the Pulp Mills case. Bolivia invokes Article 27 of
the Statute of the River Uruguay, the treaty involved in the Pulp Mills case.95
Unfortunately, however, Bolivia seems to have misread the article. Bolivia states:
“Article 27 . . . permits State Parties to use the river’s water within
their respective jurisdiction for permissible purposes without the
obligation of complying with certain procedural requirements
found in earlier provisions of the Statute, even ‘when the use is
liable to affect the regime of the river or the quality of its
waters.’”96
2.41 What Article 27 actually states is the following:
“The right of each Party to use the waters of the river, within its
jurisdiction, for domestic, sanitary, industrial and agricultural
purposes shall be exercised without prejudice to the application of
the procedure laid down in articles 7 to 12 when the use is liable to
affect the regime of the river or the quality of its waters.” 97
2.42 Thus, contrary to what Bolivia suggests, Article 27 preserves the
procedure for notification of the Commission established by the treaty. It does not
permit the parties to use Uruguay River waters within their respective jurisdictions
without having to comply with the notification procedure.
2.43 In fact, quite the contrary, according to the Court:
“Regarding Article 27, it is the view of the Court that its
formulation reflects not only the need to reconcile the varied
interests of riparian States in a transboundary context and in
particular in the use of a shared natural resource, but also the need
to strike a balance between the use of the waters and the protection
95 BR, para. 37.
96 Ibid.
97 Statute of River Uruguay, Article 27, 1295 UNTS, No. 1-21425, p. 343.
35
28
of the river consistent with the objective of sustainable
development.98
2.44 Moreover, despite the treaty-based nature of the right of usage in this
case, the Court elaborates as follows on the effect of its interpretation of Article
27 in a sentence following the one quoted by Bolivia:
“The Court wishes to add that such utilization could not be
considered to be equitable and reasonable if the interests of the
other riparian State in the shared resource and the environmental
protection of the latter were not taken into account.”99
2.45 This would presumably exclude the kind of “balancing” Bolivia seems
to have in mind, in which a State’s “rights […] to pursue its own interests and
priorities, such as development […],” could prevail over its no-harm obligation.100
2.46 The Court then concludes as follows with respect to Article 27:
“Consequently, it is the opinion of the Court that Article 27
embodies this interconnectedness between equitable and
reasonable utilization of a shared resource and the balance between
economic development and environmental protection that is the
essence of sustainable development.”101
2.47 Although Bolivia quotes most of the latter passage,102 the quote
actually confirms that Article 27 is anything but the carte blanche that Bolivia
portrays it to be. Therefore, Article 27 of the Statute of the River Uruguay, if
anything, supports Chile’s case by requiring that normal procedures of prior
notification and consultation be followed when a party, here Bolivia, uses the
98 Pulp Mills on the River Uruguay (Argentina v. Uruguay), Judgment, ICJ Reports 2010, p. 14,
p. 74, para. 177.
99 Ibid., p. 75, para. 177.
100 BR, para. 36.
101 Pulp Mills on the River Uruguay (Argentina v. Uruguay), Judgment, ICJ Reports 2010, p. 14,
p. 75, para. 177.
102 BR, para. 37.
36
29
waters of the river, “within its jurisdiction, for domestic, sanitary, industrial and
agricultural purposes […].”103
2.48 A more charitable interpretation of Bolivia’s proposed “balancing”
exercise might be that the obligation not to cause transboundary harm must be
applied not in an absolute manner, but in such a way as to permit the “acting
State” to pursue in a reasonable way its legitimate interests. If this is just another
way of saying that the no-harm obligation is one of due diligence, which the Court
has repeatedly made clear,104 Chile has no difficulty with it.
2.49 And yet the Court’s formulation of the obligation would seem to be
considerably more exacting than Bolivia suggests. Quoting from its judgment in
Pulp Mills, the Court in the joined cases of Certain Activities and Construction of
a Road, stated:
“ ‘the principle of prevention, as a customary rule, has its origins in
the due diligence that is required of a State in its territory. It is
“every State’s obligation not to allow knowingly its territory to be
used for acts contrary to the rights of other States” (Corfu Channel
(United Kingdom v. Albania), Merits, Judgment, I.C.J. Reports
1949, p. 22). A State is thus obliged to use all the means at its
disposal in order to avoid activities which take place in its territory,
or in any area under its jurisdiction, causing significant damage to
the environment of another State.’(Judgment, I.C.J. Reports 2010
(I), pp. 55‑56, para. 101.)”105
2.50 Thus Bolivia has an obligation under customary international law to
“use all the means at its disposal” to avoid activities in relation to the Silala River
103 Statute of River Uruguay, Article 27, 1295 UNTS, No. 1-21425, p. 343.
104 See, e.g., Pulp Mills on the River Uruguay (Argentina v. Uruguay), Judgment, ICJ Reports
2010, p. 14, at pp. 55-56, para. 101, and p. 79, para. 197.
105 Case concerning Certain Activities Carried Out by Nicaragua in the Border Area (Costa Rica
v. Nicaragua) and Case concerning Construction of a Road in Costa Rica Along the San Juan
River (Nicaragua v. Costa Rica), Judgment, ICJ Reports 2015, p. 665, at p. 706, para. 104.
37
30
causing significant harm to Chile. This is not an absolute obligation of result, but
one of due diligence. However, it is clear that it would not be permissible for a
State to negate the obligation by invoking “its own interests and priorities, such as
development and environmental protection and restoration”106 as a justification
for causing transboundary harm.
2.51 Bolivia concludes the section by stating that: “Whether Bolivia
decides to remove the drainage mechanisms and artificial channels, to utilize
Silala water for domestic or economic activities, or to take other action related to
the Silala within its borders lies within Bolivia’s sovereign rights.”107
2.52 The decision to do these things unquestionably lies within Bolivia’s
sovereign rights. However, as has been seen, Bolivia’s decisional sovereignty
with respect to matters within its borders does not insulate it from responsibility
for breach of the obligation not to cause significant transboundary harm. And, in
addition, Bolivia’s decisional sovereignty does not exempt it from the Court’s
declaration that a State’s utilization of shared waters within its jurisdiction “could
not be considered to be equitable and reasonable if the interests of the other
riparian State in the shared resource and the environmental protection of the latter
were not taken into account.”108 Thus, if and when Bolivia dismantles the
infrastructure in its territory, it must do so in a manner that complies with its
obligations under customary international law, including its obligation not to
cause significant harm to Chile.
106 BR, para. 36.
107 BR, para. 38.
108 Pulp Mills on the River Uruguay (Argentina v. Uruguay), Judgment, ICJ Reports 2010, p. 14,
p. 75, para. 177.
38
31
B. Bolivia continues to ignore key historical facts, most importantly its
own authorization of the channels in Bolivian territory
2.53 In its Reply, Chile pointed out four key facts that further undermine
Bolivia’s Counter-Claims and that thus far have been ignored by Bolivia:
(a) Bolivia’s century-long practice recognising the Silala as a
transboundary watercourse without distinguishing between “natural”
and “artificial” flow;
(b) The fact that the waters of the Silala were licensed in Chile to British
company FCAB in 1906, prior to the 1908 Bolivian concession and
the 1928 excavation of the FCAB channels, showing that the waters of
the Silala flowing into Chile were capable of exploitation in Chile
without the channels in Bolivian territory;
(c) The fact that the 1928 channels were constructed in Bolivia by FCAB
with Bolivian authorization, and are therefore a consequence of
Bolivia’s own sovereign acts, and
(d) Bolivia’s decision not to remove the channels and restore the wetlands
in its territory, as it could have done after the termination of the 1908
Bolivian concession in 1997.109
2.54 In its Rejoinder, Bolivia says nothing about its century-long silence on
the alleged existence of “artificial water”. As to the remaining three issues,
Bolivia either distorts historical fact or simply continues to ignore their legal
relevance.
109 CR, para. 2.52.
39
32
1. Bolivia wrongly suggests that the 1928 channels already existed at the
time of the 1906 Chilean concession
2.55 With respect to the 1928 channels, Bolivia suggests that these already
existed at the time of the 1906 Chilean concession, together with two intakes
(dams), one in Bolivian and one in Chilean territory, presumably citing from the
1906 concession as registered on the website of the Chilean Direction-General of
Water (DGA).110
2.56 Bolivia is wrong and its reference to the entry on the DGA website is
misleading. The DGA website does not describe the infrastructure as existing at
the time of the 1906 Chilean concession, but as it existed in 1990, when FCAB’s
water rights under the 1906 Chilean concession (which did not contain a volume
restriction) were adjusted to the modern Chilean Water Code and fixed at
237 l/s.111 Moreover, Chile has demonstrated that the first intake and pipeline
were constructed in 1910 and that the earth channels were excavated years after
that, in 1928, for sanitary reasons.112 Bolivia has not proven otherwise and the
entry on the DGA website does not support its case.
2. Bolivia cannot create obligations for Chile, including the obligation to
negotiate “delivery” and compensation of an alleged “artificial flow”, by
its own actions and/or omissions in relation to the channels in its territory
2.57 Bolivia maintains silence on the fact that it authorized the channels in
Bolivian territory under the 1908 Bolivian concession to the British company
110 BR, para. 48, with reference to the website of Chile’s Direction General of Water (DGA), 2019.
BR, Vol. 5, Annex 28.
111 See DGA website entry, BR, Vol. 5, p. 162.
112 CM, paras. 2.22 and 2.25; see also CR, paras. 2.61-2.64. Bolivia’s assertion that the channels’
sanitary purposes were a mere pretext (BR, para. 43), is gratuitous and has no basis in any historic
documentation.
40
33
FCAB.113 This is surprising, because Bolivia attaches important legal
consequences to the existence of the channels in Bolivia, claiming sovereign
rights on the “artificial” flow allegedly “engineered” or “produced” by the
channels in Bolivia (Counter-Claim b)), and corresponding legal obligations for
Chile related to the (albeit only future)114 delivery of this so-called “artificial
flow” (Counter-Claim c)).
2.58 Bolivia never convincingly explains why the unilateral installation of
infrastructure that regulates the Silala River in its territory should give it rights
against Chile on the optimized flow of said watercourse, when Chile played no
part in the decision to install that infrastructure or in the implementation of that
decision. Following the accepted general principles of res inter alios acta and
pacta tertiis nec nocent nec prosunt, legal doctrine is unanimous that “a State can
act unilaterally in exercise of its sovereign rights in order to reaffirm its rights, but
not in order to acquire new rights by imposing obligations on third parties without
the latter’s consent.”115
2.59 Bolivia’s default position seems to be that it is entitled to a share of
the benefits that, according to Bolivia, Chile is accruing from the channels in
Bolivia. In particular, Bolivia claims that Chile has benefitted from receiving
additional surface flow that otherwise would have entered Chile as groundwater
113 Nor does Bolivia refer to Note N° GMI-656/99 from the Ministry of Foreign Affairs of Bolivia
to the General Consulate of Chile of 3 September 1999, in which it stated that the 1908 Bolivian
concession was granted to a private company (FCAB) and not to the Chilean State and that all
actions undertaken by that company were in the private sphere and with full acknowledgement of
the Bolivian jurisdiction. See for Chile’s discussion of this point, CR, para. 2.70. The Note was
submitted by Chile as CM, Vol. 2, Annex 27.
114 See BR, para. 100: “Contrary to Chile’s assertion, Counter-Claim c) relates to the conclusion of
an agreement between the parties on the conditions and modalities of future delivery of artificiallyflowing
Silala waters from Bolivia to Chile.”
115 Third report on unilateral acts of States, by Mr. Victor Rodríguez Cedeño, Special Rapporteur,
Document A/CN.4/505 of 17 February 2000, para. 54.
41
34
and would have been more costly to exploit.116 Chile notes that numerous springs
in Chile contribute to the surface flow along the course of the Silala River and that
some of the “additional groundwater” may emerge naturally. Moreover, even if it
were the case that Chile benefitted from the channels (Chile’s experts do not
confirm a relevant increase in surface flow as a result thereof), this is still a de
facto situation that resulted from Bolivia’s own unilateral acts. It cannot constitute
the basis for rights and obligations between Bolivia and Chile vis-à-vis the waters
of the Silala, different from those arising from customary international law
regarding international watercourses.
2.60 In this context, Bolivia also claims that it is under no obligation to
maintain the infrastructure in Bolivian territory, absent any agreement between
Bolivia and Chile on the benefit accruing therefrom.117 Chile has already
confirmed in its Reply that it does not disagree with Bolivia’s (unremarkable)
statement that a State has no right to require another State to install or maintain
infrastructure for its benefit, but that this is irrelevant since Chile never requested
Bolivia to install the channels in the first place.118 Whether or not Chile obtained
any benefits from the channels as a matter of fact, is equally irrelevant. It cannot
be the source of an obligation for Chile to agree on “formulas for mutual benefit”,
as Bolivia seems to suggest.119
3. Any damage caused to the wetlands in Bolivian territory is attributable to
Bolivia, not Chile
2.61 Bolivia dedicates a significant section of its Rejoinder (pp. 22-31) to
describe the magnitude and characteristics of the channels in Bolivian territory.
116 BR, para. 83.
117 BR, para. 84.
118 CR, para. 2.49.
119 BR, para. 85.
42
35
The declared purpose of this section is to illustrate the impact that the waterworks,
if only due to their extension, must have had on the natural flow of the Silala
River.120
2.62 To prove its point, Bolivia submits with its Rejoinder two studies of
evaluation of environmental impacts in the Silala, by Bolivian non-governmental
organization FUNDECO.121 FUNDECO estimates the current extension of the
wetlands at 0.76 hectare,122 in the same range as the 0.6 hectare estimation of the
Ramsar Advisory Mission report, previously submitted by Bolivia.123 Chile notes
that DHI agrees with Chile that the Ramsar low estimate of 0.6 hectare is not
credible and that “it seems that the areas in the Ramsar report are not reflecting
the full wetland”.124
2.63 A methodological flaw common to the FUNDECO and the Ramsar
reports is the absence of a natural ecological baseline of the wetlands, prior to the
excavation of the 1928 channels. This limitation was acknowledged in the Castel
study, commissioned (but not submitted) by Bolivia, which refrained from
estimating the wetland extension prior to the channelization in the absence of
evidence.125 Chile notes that the Castel study does not support Bolivia’s thesis of
progressive degradation.126
120 BR, para. 47.
121 FUNDECO, Study of Evaluation of Environmental Impacts in the Silala, May 2018 (henceforth
“FUNDECO Study of Evaluation”). BR, Vol. 3, Annex 23.3. FUNDECO, Study of Evaluation of
Environmental Impacts in the Silala, Palynology, 2018 (henceforth “FUNDECO Study of
Evaluation Palynology”). BR, Vol. 3, Annex 23.4.
122 FUNDECO Study of Evaluation, pp. 7 and 55. BR, Vol. 3, pp. 13 and 61.
123 Ramsar Convention Secretariat, Report Ramsar Advisory Mission No 84, Ramsar Site Los
Lípez, Bolivia, 2018. BCM, Vol. 5, Annex 18.
124 DHI, Analysis and assessment of Chile’s reply to Bolivia’s counter claims on the Silala Case,
March 2019 (henceforth “DHI Analysis of Chile’s Reply”, submitted as BR, Vol.5, Annex 24),
p. 35. BR, Vol. 5, p. 41.
125 Ana Paola Castel, Multi-Temporal Analysis through Satellite Images of the High Andean
Wetlands (bofedales) of the Silala Springs, Potosí – Bolivia, September 2017 (henceforth
43
36
2.64 The Castel study points out significant inter-annual and seasonal
fluctuations in the wetlands extension.127 This factor is not taken into account in
the FUNDECO studies. According to Bolivia’s expert DHI, FUNDECO’s failure
to consider the difference between dry season and wet season vegetation, “might
cause the wrong interpretation of the data”.128 DHI also questions the
methodology used by FUNDECO to estimate the extension of the vegetation
cover of the wetlands.129
2.65 FUNDECO attributes the degradation of the wetlands directly to the
channelization in Bolivia: “This period of desiccation began around 1908, which
is a clear sign of the effects that canalization had on the Silala springs”.130 Chile
notes that FUNDECO is apparently unaware that the FCAB channels were
excavated in 1928, twenty years after the 1908 Bolivian concession, and therefore
could not explain an alleged “period of desiccation [that] began around 1908”. 131
2.66 FUNDECO’s single-cause theory for the degradation of the wetlands
is not shared by Bolivia’s expert DHI. Indeed, DHI notes that FUNDECO fails to
take into account other possible factors for desiccation, such as potential climatic
changes during the last 100-120 years, which it says may have contributed to the
“Castel”), p. 38. CR, Vol. 2, p. 271. The Castel study was cited and referenced in the DHI Report
(2018). Chile requested the Castel study, together with all other documents referred to or relied on
in Annexes 17 and 18 of the Counter-Memorial that are not publicly available, in a letter to the
Agent of Bolivia of 5 November 2018. A first set of documents was submitted by Bolivia on
22 November 2018, including the Castel study. Chile submitted a full translation of the Castel
study as CR, Vol. 2, Annex 98.
126 CR, para. 3.42.
127 Castel, p. 38. CR, Vol. 2, p. 271.
128 DHI, Technical Analysis and Independent Validation Opinion of Supplementary Technical
Studies concerning the Silala Springs, December 2018 (henceforth “DHI Technical Analysis”,
submitted as BR, Vol. 2, Annex 23), p. 22. BR, Vol. 2, p. 96.
129 DHI Technical Analysis, p. 34: “All analysed cores are taken from the central part of the
valleys and cannot be used to estimate the extension of the vegetation cover for the bofedales.”
BR, Vol. 2, p. 108.
130 FUNDECO Study of Evaluation Palynology. BR, Vol. 3, p. 142.
131 Ibid.
44
37
observed changes in the wetlands.132 DHI also suggests that the more advanced
desiccation process in the 1950s and 1960s reported by FUNDECO “can be taken
as an indicator that natural variation may also be a substantial part of the reason
for the observed changes”.133 Chile’s experts point out that cattle grazing may also
be a relevant factor, based on interviews with the local community reported
by FUNDECO.134 Chile’s experts conclude that “causal association with
channelization of the wetlands has not been proved and statements to that effect
by Bolivia and its experts are simply untrue”. 135
2.67 In any event, it is within Bolivia’s sovereign power to take the
restoration measures in its territory that it deems necessary to ensure the
biodiversity and good health of the Orientales and Cajones wetlands. For reasons
never explained, Bolivia has failed to do so thus far, even though it could have
after the termination of the 1908 Bolivian concession in 1997.136
2.68 Chile cannot of course be fixed with responsibility for Bolivia’s
inaction in its own territory. Chile encourages Bolivia to take such measures as it
deems necessary and appropriate in relation to the wetlands, whilst complying
with its obligations to Chile under customary international law, including
notification and consultation. In this context, Chile is of course willing to
coordinate with Bolivia and support its policies and regulations concerning the
132 DHI Technical Analysis, p. 25. BR, Vol. 2, p. 99.
133 DHI Technical Analysis, p. 29. BR, Vol. 2, p. 103.
134 Wheater and Peach (2019b), p. 62.
135 Ibid.
136 As Chile pointed out in its Reply, in 2000 Bolivia preferred to grant a concession to the waters
to Bolivian company DUCTEC S.R.L., for commercial exploitation to Chilean end-users. After
this scheme failed, Bolivia considered, and partly developed, several projects including a fish
farm, a small dam and a mineral water bottling plant, none of which prevailed. See CR, para. 2.75
for further details.
45
38
conservation of the wetlands in Bolivia, as established in Article 5 of the Ramsar
Convention.137
C. Conclusion: Bolivia’s claim to sovereign rights over the “artificially
enhanced flow” of the Silala River is untenable under international law and
Bolivia’s second and third Counter-Claims must be dismissed
2.69 Bolivia’s Rejoinder insists on the existence of an “artificially
enhanced” flow of the Silala River over which it claims sovereign rights. This is
the basic premise that underlies its Counter-Claims b) and c).
2.70 Leaving aside much of the jurisprudence and doctrine relied on in its
Counter-Memorial, Bolivia now claims that its sovereignty over the infrastructure
in its territory affords its sovereignty of the water allegedly “produced” thereby.
This is not an accepted theory under customary international law, which
establishes the principle of equitable and reasonable use of all transboundary
waters and the obligation not to cause significant harm. Bolivia also claims
entitlement to a share of the benefits allegedly enjoyed by Chile from the
“artificially enhanced flow”, even though the channels in Bolivia were authorized
by Bolivia without any involvement from Chile. Again, this runs counter to the
basic principle of international law that States cannot create rights for themselves
by unilaterally imposing obligations on other States. Bolivia’s Counter-Claims b)
and c) must be rejected.
137 As requested by Bolivia, see BR, paras. 95-96.
46
39
CHAPTER 3
FURTHER ANALYSIS OF THE DHI NEAR FIELD MODELLING
RECONFIRMS THAT BOLIVIA’S CONTENTIONS ON THE ALLEGED
IMPACT OF THE CHANNELIZATION ARE UNTENABLE AS A
MATTER OF FACT
3.1 In their report annexed to Chile’s Reply, Chile’s experts Wheater and
Peach demonstrated that the DHI modelling results, estimating an impact of the
channelization on surface flows of 30-40%, grossly overestimated those effects
and were based on a fundamentally flawed numerical model.138 At that time, the
DHI modelling data that had been requested by Chile had not yet been received.
3.2 The DHI modelling data made available to Chile in February 2019
confirm the conclusions reached by Chile’s experts. Moreover, their in-depth
analysis has revealed many unreported differences between the models used for
the three scenarios, which render their results incomparable. Based on this new
information, Chile’s experts have developed their assessment of the DHI
modelling. They now confirm that the DHI results “are wholly unreliable and
should be disregarded by the Court.”139
3.3 In section 3.A, Chile demonstrates that Bolivia’s response in the
Rejoinder to Chile’s criticism of the DHI Near Field Model is either insufficient
or incorrect. In section 3.B, Chile presents the most salient new insights into the
DHI modelling exercise, based on Chile’s experts’ analysis of the DHI modelling
data. While Chile recalls that Bolivia’s Counter-Claims b) and c) can and should
be dismissed solely on the legal grounds set out in Chapter 2, the basic errors
made by Bolivia’s expert DHI further undermine the credibility of Bolivia’s
Counter-Claims b) and c).
138 Wheater and Peach (2019a). CR, Vol. 1.
139 Wheater and Peach (2019b), pp. 9 and 64.
47
40
A. DHI’s response to the fundamental criticism of Chile’s experts of the
Near Field Model set-up is insufficient or incorrect
1. DHI acknowledges that the model results of the three scenarios (Baseline,
No Canal and Undisturbed) are most likely affected
by the Near Field boundary conditions
3.4 In their Reply report, Chile’s experts Wheater and Peach criticized
certain basic assumptions of the DHI Near Field modelling used to evaluate the
impact of the channelization in Bolivia on cross-boundary flows, which was based
on three different scenarios: the Baseline scenario, the No Canal scenario and the
Undisturbed (or Restored Wetlands) scenario.
3.5 Wheater and Peach showed that the DHI decision to model a small
area (approximately 1% of the entire Silala water catchment), drawn closely
around the man-made channels in Bolivia (the Near Field) with “fixed head”
boundary conditions, grossly exaggerates the modelled differences between the
scenarios. In particular, and remarkably, it results in very different inflows in the
three scenarios, even though the recharge area and precipitation remain the
same.140 The different inflows in the three scenarios are the main driver of the
different outflows in the scenarios. Thus, the alleged 30-40% “enhanced flow” is a
result of the modelling exercise, not of the channelization.141
3.6 DHI now acknowledges the impact of its modelling set-up on the
results obtained. Specifically, DHI recognizes that “in the Near Field model
140 A “fixed head” boundary condition allows the model to draw on an infinite amount of water,
which means that when the hydraulic gradient within the model area changes water will flow into
the model in response to the changed hydraulic gradient, irrespective of whether that water is
available in the natural system. See Wheater and Peach (2019a), pp. 20-21. CR, Vol. 1, pp. 116-
117. It goes without saying that no infinite amount of water is available in the Atacama Desert.
141 Wheater and Peach (2019a), pp. 5-6 and 44. CR, Vol. 1, pp. 101-102 and 140.
48
41
groundwater flows are closely tied to groundwater boundary conditions. DHI
acknowledges that the model results, including the impacts of the drainage
network, are most likely sensitive to boundary conditions”.142 DHI’s conclusion is
that their previous results represent an upper bound estimate of what might be
feasible effects,143 which suggests that – even according to DHI – these earlier
results are not a realistic assessment.144
3.7 DHI also acknowledges that reduced surface flow will be
compensated by increased groundwater flow145 (except for the increased
evapotranspiration). Thus, DHI does not dispute that the overall effect of the
channels in Bolivia on the total cross-boundary flow (surface and groundwater) is
practically non-existent.146 Insofar as the scientific and technical facts are relevant
at all, this is the one fact that the Court needs to focus on. The Silala waters flow
as surface water or as groundwater, down the hydraulic gradient, into Chile,
regardless of the channelization.
3.8 Returning to the issue of the quantum of specifically surface flows, in
response to Chile’s criticisms, DHI conducts a sensitivity analysis of the Near
Field Model, resulting in a significant reduction of its earlier estimate of the
142 DHI Analysis of Chile’s Reply, p. 23. BR, Vol. 5, p. 29. They also recognize that “[w]ith
higher groundwater heads internally in the Near Field model less water enters the model domain
through the fixed head boundaries as the gradient changes”. DHI Analysis of Chile’s Reply, p. 24.
BR, Vol. 5, p. 30.
143 DHI Sensitivity Analysis, p. 7. BR, Vol. 5, p. 55.
144 Wheater and Peach (2019b), p. 20.
145 DHI Analysis of Chile’s Reply, p. 24: With no canals, less waters enters the surface water
system and more enters the groundwater inside or outside the Near Field model domain. […]
Groundwater entering the Near Field or flowing past the Near Field domain will likely flow into
Chile.” BR, Vol. 5, p. 30.
146 BR, para. 68 points to a difference in flow velocity between surface and groundwater flow,
apparently suggesting that this would affect the availability of water in Chile. Bolivia’s own
experts DHI disagree: “Differences in flow velocities between a porous media and a canal is not an
appropriate measure of change or modification.” DHI, Technical Analysis, p. 13. BR, Vol. 2,
p. 87.Chile’s experts explain that the flow velocity does not affect the discharge across the border.
Wheater and Peach (2019b), p. 1, footnote 1.
49
42
impact of the channels on the surface flow, from 30-40% to 11-33%.147 As noted
above, in these new simulations, the “fixed head” boundary assumption provides
the upper bound results.148 The lower bound results are obtained by assuming no
change of groundwater inflow into the Near Field Model area in the different
scenarios, i.e. by replacing the “fixed head” boundary condition by a “fixed flux”
boundary condition.149 Chile’s experts agree that this is an appropriate
assumption.150 The lower bound results are contained in Table 1 below:
Table 1. DHI Sensitivity Analysis, p. 32, Table 7-2. BR, Vol. 5, p. 80.
3.9 Notwithstanding the adjustment of the “fixed head” boundary
condition, the DHI lower bound results in further difficulties.
3.10 The principle of conservation of mass dictates that, for a steady-state
simulation, as specified by DHI,151 inflow must equal outflow. However, DHI’s
lower bound simulation of the impact of channelization shows a decrease of 16 l/s
in surface outflow, an increase of 4 l/s in groundwater outflow and an increase in
evapotranspiration of 3 l/s in the No Canal scenario, as compared to the Baseline
scenario, leaving 9 l/s unaccounted for in the No Canal scenario. This makes no
147 DHI Sensitivity Analysis, p. 8. BR, Vol. 5, p. 56.
148 DHI Sensitivity Analysis, p. 7. BR, Vol. 5, p. 55.
149 DHI Sensitivity Analysis, p. 31. BR, Vol. 5, p. 79.
150 Wheater and Peach (2019b), p. 9.
151 DHI Report (2018). BCM, Vol. 5, p. 67.
50
Table 7-2 Results of the outer bounds of the sensitivity analyses of the upper head boundary conditions as changes
in 1/s from the flow components in the baseline simulation with the canals
Canalised
Changes from canalised conditions (1/s)
situation (1/s)
Baseli ne Lower Bound Upper Bound
Inflow to model 253.6 -1 -27.9
Surface outflow 149.0 -16 -48.6
Groundwater outflow 106.3 4 10.8
Evapotranspi ration 10.0 3 3.4
Storage and num. inacuracy -11.7 8.4 6.6
43
sense. It is due to the change in storage and numerical inaccuracy of the lower
bound results, stated by DHI to be 8.4 l/s, and is more than half of the supposed
impact of the channels the model is intended to estimate. Further, DHI’s stated
error in the Baseline simulation is 11.7 l/s. This means that the combined model
errors exceed the estimated effect of the channels, discrediting the modelling
results.152
3.11 The missing water and large margin of error in the DHI sensitivity
analysis are indicative of major issues with the DHI model set up, beyond the
“fixed head” boundary issue noted by Chile’s experts in their Reply reports.153
These additional issues have been identified, thanks to the modelling data
submitted by Bolivia in February 2019, and will be addressed in section 3.B
below.
2. DHI’s characterization of Chile’s experts analysis as “vastly simplified”
is misplaced
3.12 DHI considers that Chile’s criticisms of the DHI modelling are
“highly simplified and ignore the peculiarities of the flow of Silala waters.”154 In
particular, DHI questions Chile’s “simplified impact calculations” that, according
to DHI, “do not support the claim that DHI’s impacts are exaggerated. The
analysis is based on the one-dimensional Darcy equation, which is only valid
under idealized conditions not satisfied at Silala.”155
3.13 DHI refers to Chile’s text-book calculation of hillslope groundwater
flows (Appendix 1 to the Wheater and Peach (2019a) report), but misinterprets its
152 Wheater and Peach (2019b), pp. 21-22.
153 Wheater and Peach (2019b), p. 22.
154 BR, para. 60.
155 DHI Analysis of Chile’s Reply, p. 7. BR, Vol. 5, p. 13.
51
44
purpose. This text-book calculation was used by Chile’s experts to demonstrate
the effect of a “fixed head” boundary condition close to the modelled channels, on
the Near Field modelling results. It was never intended, nor used, to produce
Chile’s own estimate of the expected effect of the channelization.156 Chile’s
experts would be the first to acknowledge that the calculations are simplified, but
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.”157
They stand by this conclusion and find further support in the modelling data as
will be set forth in section 3.B.
3.14 Bolivia and its experts also question Chile’s “simplified analysis” of
the complex geology of the Silala River area and signal that this “is a clear
inconsistency, which brings into question the validity of their assessments of the
canalization impact”.158
3.15 In so far as this comment is directed at Chile’s experts’ text-book
calculation of the exaggeration factor of DHI’s modelling results, it is misdirected.
As said above, this calculation does not purport to assess the impact of the
channelization impact and does not consider, nor intend to consider, the real
geology or geography of the Silala River area.
3.16 In so far as the comment is directed at Chile’s substantive criticism of
the geology and hydrogeology of the Silala River basin on which the DHI model
is built, it is unfounded. Far from a “simplified analysis”, Chile’s experts have
thoroughly reviewed the evidence provided by Bolivia’s experts and found it
severely wanting.
156 Wheater and Peach (2019b), p. 4.
157 Wheater and Peach (2019a), Appendix 1, p. 58. CR, Vol. 1, p. 154.
158 BR, para. 61.
52
45
3.17 For example, Bolivia has insisted on the existence of an alleged fault
system with high hydraulic conductivities, running from the Orientales wetland to
the Cajones wetland, bending around to follow the course of the Silala River from
Bolivia into Chile.159 This so-called “Silala Fault” is provided by Bolivia as an
alternative for the origin, location and alignment of the Silala River ravine, which
they say was later eroded by glacial and fluvioglacial action rather than the purely
fluvial origin of the ravine, which was demonstrated by Chile.160 As Chile’s
experts pointed out, no evidence was provided with the Bolivian Counter-
Memorial to support the existence of this major fault.161 In the words of Chile’s
experts, its existence is “so unlikely that we believe it impossible”.162
3.18 In the Bolivian Rejoinder, Bolivia submits additional studies by
Bolivian expert F. Urquidi,163 the Bolivian geological survey SERGEOMIN,164
and the Tomás Frías Autonomous University of Potosí (TFAU),165 to support the
existence of the “Silala Fault” and associated fractures.166 Chile’s experts have not
159 Peach, D.W. and Wheater, H.S., Concerning the Geology, Hydrogeology and Hydrochemistry
of the Silala River Basin, 2019 (henceforth “Peach and Wheater (2019)”), p. 29. CR, Vol. 1, 195.
160 Wheater and Peach (2019b), p. 50.
161 Peach and Wheater (2019), p. 32. CR, Vol. 1, p. 198.
162 Peach and Wheater (2019), p. 23. CR, Vol. 1, p. 189.
163 BR, Vol. 3, Annex 23.5 (henceforth “Urquidi Report”). Chile notes that the Urquidi Report is
particularly unfavourably reviewed by Bolivia’s own expert DHI, in the DHI Technical Analysis
Report. BR, Vol. 2, Annex 23. DHI points out inconsistencies in data used and conclusions drawn,
which contradict DHI’s own findings (BR, Vol. 2, p. 79). Some of the data relied on in the report
is confirmed to be of dubious quality, and several statements are found to be unsupported by data
and therefore cannot be reasonably evaluated (BR, Vol. 2, p. 115). DHI even questions the
author’s objectivity: “[T]he reviewer recommends limiting the findings to technical determinations
rather than non-technical conclusions such as whether Silala is an international river course. […]
However, the author’s narrative seems not solely based on technical objective interpretation of
data and analyses presented. This impression stems from the numerous unsupported statements
and lack of presentation of or referencing materials on which the author’s conclusions are drawn.”
(BR, Vol. 2, pp. 115-116).
164 BR, Vol. 3, Annex 23.5, Appendix a (SERGEOMIN 2003); BR, Vol. 4, Annex 23.5,
Appendix b (SERGEOMIN 2017).
165 BR, Vol. 3, Annex 23.5, Appendix c (TFAU 2018).
166 BR, para. 62.
53
46
found any evidence for the existence of a Silala fault in these reports. DHI relies
on the SERGEOMIN (2017) geologic mapping that “indicates a relatively small
displacement of 5 m. at the border”, as evidence of the fault.167 However, Chile’s
experts confirm that any displacement by the alleged fault as proposed by Bolivia
would be in the opposite direction than indicated by the Bolivian geologists and
that the 5 m displacement is more reasonably explained by the fact that the Silala
ignimbrite in Bolivia is tilted to the west.168
3.19 More generally, Chile’s experts find that the new reports submitted by
Bolivia “contain important errors and inconsistencies in the understanding and
interpretation of stratigraphy, structural geology and in petrography. These errors
and inconsistencies have led to errors in development of a hydrogeological
conceptual model, upon which DHI relied for the design of their numerical
models, particularly, the NFM [Near Field Model].”169
3.20 In particular, Chile’s experts note significant confusion in the Bolivian
reports concerning the ages of the various geological units.170 Bolivia’s geologists
also invert the sequence of rock strata (or stratigraphy), mapping older geological
units as overlying younger geological units, which is clearly impossible.171 This
means that the Near Field Model is based on an incorrect understanding of the
geometry and extent of the deep aquifer in the permeable ignimbrite deposits,
including the lack of inclusion of a shallower aquifer found in Chile.172 This has
considerable consequences for the modelled groundwater flow regime that the
167 DHI Analysis of Chile’s Reply. BR, Vol. 5, p. 24.
168 Wheater and Peach (2019b), p. 50.
169 Wheater and Peach (2019b), p. 43.
170 Wheater and Peach (2019b), pp. 44-45.
171 Wheater and Peach (2019b), pp. 46-49.
172 Wheater and Peach (2019b), p. 56.
54
47
DHI models are purporting to represent.173 This is compounded by inconsistencies
and errors in the naming of rock types, leading to an inability to have confidence
in the Bolivian geological interpretations.174
3.21 Chile’s experts point out that DHI acknowledges the existence of two
distinct sources of groundwater discharging to the Silala springs, but has decided
not to include this in the model, because “it would not reflect on the split between
surface and groundwater discharge from the wetlands due to channelization”.175
Chile’s experts disagree. Groundwater flows from different aquifers with different
piezometric heads will interact differently with surface water. Not taking this into
account will lead to an incorrect understanding of the groundwater flow regime,
and most likely incorrect modelling results.176
3.22 Chile’s experts conclude that the new evidence presented by Bolivia,
in response to Chile’s criticism of the DHI modelling, confirms their view that the
geology and hydrogeology that underpin the DHI’s modelling is incorrect and
does not represent reality. If only for that reason (there are many others), the
models are likely to produce results that are very seriously in error.177
173 Wheater and Peach (2019b), p. 56.
174 Wheater and Peach (2019b), pp. 53-54.
175 DHI Analysis of Chile’s Reply, p. 21. BR, Vol. 5, p. 27.
176 Wheater and Peach (2019b), pp. 54-55.
177 Wheater and Peach (2019b), p. 11.
55
48
B. The DHI modelling data submitted on 7 February 2019 reveal
further flaws in the Near Field Model which render the modelling results
wholly unreliable
1. DHI uses different models for the Baseline Scenario on the one hand, and
the No Canal and Undisturbed Scenarios on the other hand, which make
them not directly comparable
3.23 The modelling data received in February 2019 revealed that DHI uses
a different configuration of models for the Baseline Scenario on the one hand, and
the No Canal and Undisturbed (or Restored Wetlands) Scenarios on the other
hand, all within the Near Field Model area.
3.24 In all three scenarios, DHI uses the MIKE-SHE model to represent the
hydrology of the Silala River basin, including coupled surface and groundwater
flows, evapotranspiration, unsaturated zone and overland flow.178 In addition, but
only in the Baseline Scenario, DHI uses the MIKE-11 model to represent the
channel flow. MIKE-SHE and MIKE-11 are coupled in the Baseline Scenario,
which allows exchanges of water between the channels and the groundwater
system.179 In the No Canal and Undisturbed scenarios, the channels are not
explicitly modelled and so the MIKE-11 model is not included. This has the effect
of directing all surface flow in these two scenarios as overland flow, as if a river
does not run through its own natural channels.180 This is of course incorrect. The
fact that a different modelling framework is used for the Baseline Scenario is one
178 Wheater and Peach (2019b), p. 17.
179 Wheater and Peach (2019b), pp. 17 and 23.
180 Wheater and Peach (2019b), p. 23.
56
49
reason why the three scenarios are not directly comparable.181 Other important
reasons are explained below.
2. The modelling conditions, including surface topography and inflow, are
different in each scenario, invalidating any comparison of the modelling
results
3.25 More disturbing than the use of different model configurations in the
three model scenarios is the fact that there are basic unreported differences
between the three scenarios that have only been detected by Chile’s experts after
thorough analysis of the modelling data.
3.26 Clearly, for the three scenarios to be comparable, their topography
should be identical, except for the differences resulting from the elimination of the
channels in the No Canal and Undisturbed Scenarios and an allowance for
possible peat growth in the Undisturbed Scenario.182 Chile’s experts were
therefore very surprised indeed to find that very different topographies were used
for the MIKE-SHE model for the Baseline Scenario as compared with the other
two scenarios, and that different topographies had been used in the Baseline
Scenario for the MIKE-SHE and the MIKE-11 models.183 As can be seen in
Figure 1, the topography of the No Canal and Undisturbed scenarios show
differences of almost 7 m from the Baseline MIKE-11 model and almost 3 m from
the Baseline MIKE-SHE model. This is far more than the 0.50 – 1.00 m depth of
the channels, or the 60 cm possible peat growth in the Undisturbed Scenario, the
effects of which the model scenarios are meant to simulate.184
181 Wheater and Peach (2019b), p. 23.
182 Wheater and Peach (2019b), p. 23.
183 Wheater and Peach (2019b), pp. 23-24.
184 Wheater and Peach (2019b), p. 24.
57
50
Figure 1. Ground surface elevations used in the four Bolivian models compared at two
cross sections of the main channel near the international border (Muñoz et al., 2019,
p. 30, Figure 4-3. CAP, Vol. 2, Annex XV).
3.27 The raising of the ground surface in Bolivia in the No Canal and
Undisturbed Scenarios has the effect of reducing the groundwater inflows into the
model and increasing the groundwater outflow, so less water will be available to
appear as surface flow, in comparison to the Baseline Scenario. This is exactly the
58
Cross section no. 3560
4396
4394 _
4288
439J !
4290 _~ -~.. 1- ,......,. .........., ......,. ..........., ......,. .....~ -..-----...
J
4286 t j
Distance (m)
2 4 6 8 10 12 16
4302
4300 ]:
C:
4298 ,Q "~'
Cross section no. 3370
4296 ~u-:; -.:!::==l=t B
4294
4292
Dista ce(m)
4 6 8 10 12
- Baseline (MIKE-11) - Undisturbed (MIKE-SHE) • • • No Canal (MIKE-SHE) Baseline (MIKE-SHE)
51
effect that Bolivia attributes to the removal of the channels. This renders the
whole modelling exercise meaningless.185
3.28 Another unreported and disturbing difference between the three model
scenarios concerns the inflows in each scenario. Chile’s experts had noted
unexplained differences in the volume of inflow in the three scenarios, from the
DHI report submitted with Bolivia’s Counter-Memorial.186 However, they
assumed that the inflow in all three scenarios corresponded to groundwater input
across the Near Field Model boundaries.187
3.29 It is now apparent from the modelling files that, apart from the
groundwater inflow, distinct amounts of additional water are introduced to the
Near Field Model, at the location of the springs: 42 l/s in the Baseline Scenario
and 31 l/s in the No Canal and Undisturbed Scenarios. In the Baseline Scenario,
this water is added directly to the surface flow, whereas in the No Canal and
Undisturbed Scenarios it is partitioned between surface and groundwater flow.188
The difference of 11 l/s and their different treatment as surface and groundwater
flow is arbitrary and neither reported nor explained by DHI in its reports.189 The
difference of 11 l/s is of the same order of magnitude as the purported effects of
the channels that the models are designed to simulate.
3.30 Further anomalies in the modelling are reported in more detail in the
Wheater and Peach (2019b) expert report submitted with the Additional Pleading,
based on the detailed analysis of the modelling data undertaken in the Muñoz et
al. (2019) report, also submitted hereby.190 However, the different topographies in
185 Wheater and Peach (2019b), p. 27.
186 Wheater and Peach (2019b), p. 49.
187 Wheater and Peach (2019b), p. 32.
188 Wheater and Peach (2019b), p. 33, footnote 6.
189 Wheater and Peach (2019b), p. 33.
190 Muñoz et al. (2019). CAP, Vol. 2, Annex XV.
59
52
the three scenarios, compounded by the unaccounted-for additional 11 l/s added to
the Baseline Scenario, are inexplicable errors made by DHI that render their
modelling results wholly unreliable. The DHI modelling results should be
dismissed by the Court.
3. Bolivia’s reliance on the 1922 Fox flow estimate to confirm the results
obtained from the DHI modelling is not credible
3.31 Bolivia seeks support for the DHI modelling results in field
observations conducted in 1922 by FCAB engineer Robert H. Fox, which it says
documented 131 l/s cross-border flow.191 According to Bolivia, this is 18-38%
lower than present observations (160-210 l/s), and therefore corresponds to the
11-33% findings of the DHI sensitivity analysis.192
3.32 Bolivia’s reliance on the Fox flow estimate is not credible, and it is
even misleading.
3.33 First, there is no evidence in the Fox article for Bolivia’s affirmation
that the measured daily flow of 11,300 cubic metres (or 131 l/s) corresponds to the
cross-border flow. The article does not specify the precise location of the
measurement. However, a possible inference from the article is that the flow was
measured at the “small dam” (or Intake N° 1) that was built in 1909 and operative
from 1910 onwards, located in Bolivian territory just below the confluence of the
Cajones and Orientales ravines, at approximately 600 m upstream from the
international boundary.193 Both Bolivia’s and Chile’s experts report major
changes to the flow rate at different locations along the river, due to continuous
191 BR, para. 65. Reference is to R.H. Fox, “The Waterworks Department of the Antofagasta
(Chili) & Bolivia Railway Company”, South African Journal of Science, 1922, p. 123. CM, Vol. 3,
Annex 75.
192 BR, para. 65.
193 See for the location of Intake No 1, CM, para. 2.22.
60
53
interaction between surface and groundwater flows.194 Hence, no conclusions can
be drawn about the cross-border flow from one single measurement, of uncertain
location.195
3.34 Second, the methodology used for the measurement is unknown, and
hence its accuracy and reliability.196 Chile’s and Bolivia’s experts concur that,
even in the Twenty-First century, the Silala River flow is notoriously difficult to
measure accurately, due to the extreme conditions of the altiplano.197 DHI reports
that comparison of long term flow records from the permanent gauging stations in
Bolivia and Chile “shows significant differences in both the mean flow levels and
temporal variation”.198 DHI also notes that none of the series from these two sites
seems to be free of gauging inconsistencies, “which may be due to the remote
locations and harsh climate”.199 In addition, short term measurements made by
Bolivia in 2017 show “inconsistencies both at the individual gauging points and
also when cross comparing the data”.200 Fox’s single estimate is prone to these
same limitations and more, and cannot be accepted at face value.
3.35 In short, a century-old single estimate, at a location that is uncertain,
but possibly 600 m removed from the international boundary, using an unknown
methodology, cannot credibly be used to validate the results of the DHI
modelling, which for many reasons, as explained before, have no scientific value
and should be disregarded by the Court.
194 Wheater and Peach (2019b), p. 38.
195 Wheater and Peach (2019b), p. 39.
196 Wheater and Peach (2019b), p. 39.
197 Wheater and Peach (2019b), p. 39.
198 DHI Report (2018). BCM, Vol. 2, p. 395.
199 Ibid.
200 Ibid.
61
54
C. Conclusion: The DHI Near Field modelling is so severely flawed that
its results are meaningless and should be dismissed by the Court.
3.36 The dispute before the Court can and should be decided solely on
legal grounds, and Bolivia’s Counter-Claims b) and c) must be rejected because
they have no basis in international law.
3.37 That being said, Chile has made use of its right to respond to Bolivia’s
factual allegations, in particular, that the channelization of the Silala River on
Bolivian territory has caused its cross-border surface flow to increase by an
estimated 11-33% (formerly 30-40%). Bolivia relies entirely on its expert DHI,
and the results of the DHI Near Field modelling, for this estimation. An
evaluation of the quality of the DHI modelling exercise is therefore of particular
interest to Chile and the Court.
3.38 Chile’s experts have demonstrated that the DHI Near Field modelling
is not scientifically sound, for the following reasons:
(a) completely different topographies were used for the different
scenarios, with the differences in topography very much larger than
the relatively small differences associated with channelization and
assumed peat growth;
(b) there are basic errors in the geology and hydrogeology that mean that
the geometry and aquifer properties are wrong;
(c) arbitrary amounts of water have been added to the wetland springs,
with different amounts for the different scenarios (and different
partitioning between surface and groundwater), thus adding to the
simulated effects of channelization, and
62
55
(d) errors and inaccuracies, including changes in storage, account for
similar rates of flow to the effects ascribed to the channelization and
peat growth.201
3.39 Chile’s experts have stated in plain language that they are shocked by
the basic errors made by DHI.202 The results of the DHI Near Field modelling are
unreliable, even meaningless, and should be dismissed by the Court.
201 Wheater and Peach (2019b), p. 40.
202 Wheater and Peach (2019b), p. 40.
63

57
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
16 September 2019
65

Expert Report
Wheater, H.S. and Peach, D.W., Impacts of Channelization
of the Silala River System in Bolivia on the Hydrology of the
Silala River Basin – an Updated Analysis
67

IMPACTS OF CHANNELIZATION OF THE SILALA RIVER IN
BOLIVIA ON THE HYDROLOGY OF THE SILALA RIVER BASIN –
AN UPDATED ANALYSIS
Drs. Howard Wheater and Denis Peach
August 2019
69

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
71
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 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, 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.
72
v
TABLE OF CONTENTS
LIST OF FIGURES............................................................................................... vii
LIST OF TABLES ................................................................................................. ix
LIST OF ACRONYMS AND ABBREVIATIONS................................................ x
1 INTRODUCTION......................................................................................... 1
1.1 Experts’ Terms of Reference ........................................................................ 1
1.2 Report background and structure .................................................................. 3
2 SUMMARY................................................................................................... 6
3 MODELLING THE IMPACTS OF CHANNELIZATION ON THE
HYDROLOGY OF THE SILALA RIVER SYSTEM................................ 11
3.1 Introduction........................................................................................... 11
3.2 Water balance modelling ...................................................................... 14
3.3 Near Field modelling ............................................................................ 17
3.3.1. The Pleadings to date ................................................................ 17
3.3.2. Information from DHI’s model files......................................... 22
3.3.3. Topography of the modelled scenarios ..................................... 23
3.3.4. Hydrogeology and boundary conditions................................... 27
3.3.5. Numerical performance and other modelling issues................. 31
3.3.6. Flow estimate from Fox (1922) ................................................ 38
3.4 Conclusions........................................................................................... 40
4 ISSUES RELATED TO THE GEOLOGY AND HYDROGEOLOGY
OF THE SILALA BASIN ........................................................................... 41
4.1 Introduction........................................................................................... 41
73
vi
4.2 Geology and Hydrogeology of the Silala River wetland springs and
surrounding area ................................................................................... 43
4.2.1 Stratigraphy............................................................................... 44
4.2.2 Structural Geology.................................................................... 50
4.2.3 Petrography............................................................................... 53
4.2.4 Hydrogeology ........................................................................... 54
4.3 Conclusions........................................................................................... 57
5 ERRONEOUS STATEMENTS AND FLAWED ATTRIBUTION
CONCERNING WETLAND DEGRADATION ........................................ 59
6 CONCLUSIONS ......................................................................................... 62
7 REFERENCES ............................................................................................ 67
74
vii
LIST OF FIGURES
Figure 1. Hydrological catchment and SNF area defined by DHI (BCM, Vol. 2,
p. 328).................................................................................................................... 12
Figure 2. Domains covered by the three different DHI models (Muñoz et al.,
2019)...................................................................................................................... 13
Figure 3. Boundary conditions of the Water Balance Model (Muñoz et al., 2019).
............................................................................................................................... 16
Figure 4. The area of Bolivia’s Near Field Model. The green color represents
Fixed Head boundary conditions, the black color represents No-Flow boundary
condition and the grey color represents Fixed Gradient boundary condition
(Muñoz et al., 2019). ............................................................................................. 19
Figure 5. Ground surface elevations used in the four Bolivian models compared at
two cross sections of the main channel near the international border (Muñoz et al.,
2019)...................................................................................................................... 25
Figure 6. Ground surface elevations used in the four Bolivian models compared at
two cross sections of the main channel in the Orientales wetland. Specifically, in
these cross sections the Baseline and the No Canal topographies from the MIKESHE
model coincide and the black dotted line obscures the yellow line (Muñoz et
al., 2019)................................................................................................................ 26
Figure 7. Groundwater level maps used in definition of groundwater component
boundary conditions (BCM, Vol. 5, p.19). Black lines represent the piezometric
contours, the polygons filled with grey lines represent the HGU4 unit. The added
red arrow represents the implied groundwater flow through the HGU4 unit
(Muñoz et al., 2019). ............................................................................................. 28
Figure 8. A) Groundwater level contours in the Silala NFM, interpolated from
piezometer wells, spring elevations, and wetlands excavations for soil sampling
(Adapted from BCM, Vol. 4, p. 97). The NFM domain is delimited by the
polygon with a black and white border that shows in black the DHI no-flow
boundaries and in white the boundaries through which water can pass. The blue
arrows represent the direction of groundwater flow interpreted from the contour
lines. B) Northern and C) Southern wetlands overall flow directions. (Adapted
from BCM, Vol. 2, p. 371). Note: The text in the lower label of panel C is:
“Drained part of wetland with signs of vegetation changes” (Muñoz et al., 2019).
............................................................................................................................... 30
75
viii
Figure 9. Initial potential head difference between the Baseline and No Canal
scenarios and the Undisturbed Scenario. Positive values correspond to locations
where the initial potential head is higher in the Undisturbed Scenario (Muñoz et
al., 2019)................................................................................................................ 32
Figure 10. Some of the cross sections where the water does not flow through the
main channel. (Muñoz et al. 2019)........................................................................ 35
Figure 11. Cross sections where the channel is flooded (Muñoz et al., 2019)...... 36
Figure 12. Flow variations in two sections of the reach between the Cajones and
Orientales confluence and the international border (A) Plan view of the NFM
domain that depicts the locations of the two sections analyzed. (B) Discharge time
series of the flow at sections 3550 and 3160. (C) Zoom into the last two days of
the flow at sections 3550 and 3160 (Muñoz et al., 2019). .................................... 37
Figure 13. The Silala Ignimbrite (Chi) clearly overlying Silala Ignimbrite Nis-3
(Bol), located at the East of Bofedales Sur (Orientales). The younger Inacaliri
Lavas 2 overlie the Silala Ignimbrite (Chi) (SERNAGEOMIN (Chile), 2019b).. 46
Figure 14. Geological Map of the Silala Study area (BR, Complete Copies of
Certain Annexes, Vol. 2, Annex 23.5 Appendix a, p. 69). (A) Expanded view of
the cross section and (B) legend from Bolivia’s geological map reproduced at
Figure 2-1.. ............................................................................................................ 48
Figure 15. Generalized Geological Section of the Silala Springs (BR, Vol. 4,
p. 125) shows that the ages cited by Bolivia (added for clarity in the figure) do not
support the stratigraphic relationship determined by Bolivian geologists (BR, Vol.
4, p. 115). In the cross-section, older rocks (Miocene Volcanics) are seen to
overlie younger lavas. This is clearly incorrect. The section also shows the
ignimbrites underlying the lavas of the Silala Chico (Cerrito de Silala) dome,
which is also incorrect, since they are younger than the lavas of the dome. The
Chilean age for the Silala Ignimbrite (Chi) is 1.61 Ma and for the Cabana
Ignimbrite (Chi) is 4.12 Ma (SERNAGEOMIN (Chile), 2019b) ......................... 49
Figure 16. Amended map from DHI, 2018 (BCM, Vol. 4, p. 76, Figure 29)
showing in red the postulated Silala Fault system (CR, Vol. 3, p. 211). .............. 51
Figure 17. UKFS, located 31 km to ENE of the Silala River basin groundwater
catchment. The abbreviation KUFZ in the figure refers to the UKFS, which has
been used in this report (SERNAGEOMIN (Chile), 2019b). ............................... 53
76
ix
LIST OF TABLES
Table 1. Water balance from the “Water Balance Model” -
Silala_model_gw_200m_v12_final.she version (Muñoz et al., 2019).................. 15
Table 2. Different components of the water balance of the model (BR, Vol. 5,
p. 80)...................................................................................................................... 21
77
x
LIST OF ACRONYMS AND ABBREVIATIONS
BCM – Counter-Memorial of the Plurinational State of Bolivia
BR – Rejoinder of the Plurinational State of Bolivia
CM – Memorial of the Republic of Chile
CR – Reply of the Republic of Chile
DHI – Danish Hydraulic Institute
ha – hectares
km2 – square kilometres
l/s – litres per second
m – metres
mm/year – millimetres per year
NDVI – Normalized Difference Vegetation Index
NFM – Near Field Model
UKFS – Uyuni-Khenayani Fault System
WBM – Water Balance Model
78
1
1 INTRODUCTION
1.1 Experts’ Terms of Reference
In 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, there has been considerable agreement by both
parties and their technical experts, through the exchange of written pleadings. It is
agreed (CR, Vol. 1, pp. 99-101; BCM, Vol. 2, pp. 266-267) that the Silala River is
an international watercourse, and that both surface water and groundwater from
the Silala River topographic catchment and a larger groundwater catchment
naturally flow from Bolivia to Chile. It is also agreed that the channelization of
the wetlands and natural Silala River channel in Bolivia is expected to have
increased the surface water discharge and decreased the groundwater flow from
Bolivia to Chile1. Further, it is agreed by both parties that if the channels were
removed and the wetlands restored to their pre-channel state, there would be a
reduction in surface flow, a small increase in wetland evaporation, and an increase
in cross-border groundwater flow.
An important area of continuing scientific disagreement between the two parties
concerns the magnitude of the impacts of the channelization on trans-border
surface and groundwater flows. Bolivia has based its estimates primarily on
modelling by the Danish Hydraulic Institute (DHI) of the Near Field, a small area
1 We note some confusion in Bolivia’s understanding of these changes. BR, para. 68, correctly
states that surface water flows much faster than groundwater, but seems to imply, mistakenly, that
this has relevance to the points of legal or scientific disagreement (BR, Vol. 1, p. 37). What is
important to the dispute is the distribution of water resources between surface water and
groundwater, which depends on the discharge of water, as a volumetric flow rate (l/s), and not the
flow velocity (m/s). Bolivia notes (in the same paragraph) that ‘The surface flows of the Silala
from Bolivia to Chile have been enhanced in terms of volume and flow.’ (BR, Vol. 1, p. 37). The
use of the terms ‘volume’ and ‘flow’ shows some conceptual confusion. It is the volume flow rate
that is important, and not the flow velocity. The fact that surface water flows more rapidly than
groundwater does not affect the discharge (volumetric flow rate) across the border. The high
velocity in the river channel applies to a small river cross section, whereas the orders of magnitude
slower velocity of groundwater applies to a large geological cross-section of aquifer.
79
2
including the Bolivian headwater springs and wetlands and the river corridor to
the international border. This modelling produced large estimates of the impact,
whereas in the opinion of Chile’s experts, these effects will be small.
In this context, the Republic of Chile has requested our independent expert
opinion, as follows:
“Questions for Dr. Howard Wheater, as a hydrological engineer:
(i) Do the digital data provided by Bolivia in support of the
Counter-Memorial after Chile’s Reply was finalized
materially change your assessment of the modelling by
Bolivia’s Experts of the effects of channelization and of the
effects of possible long-term peat growth?
(ii) Do the further modelling and other studies by Bolivia’s
Experts, presented in Bolivia’s Rejoinder, reflect a correct
assessment of the magnitude of these effects?
Questions for Dr. Denis Peach, as a hydrogeologist:
(i) Does the information provided by Bolivia’s Experts,
presented in Bolivia’s Rejoinder, materially change your
assessment of the geology and hydrogeology that underpins
Bolivia’s modelling of the effects of channelization and of
the effects of possible long-term peat growth?
(ii) What are the implications of this new information, if any, for
the validity of Bolivia’s modelling of the effects of
channelization and possible long-term peat growth?”
This report is restricted to addressing these questions. At this stage of the
proceedings, we feel it is a distraction to the Court to rebut all of Bolivia’s
misinterpretations of our previous reports, or to address other errors in the science
reports presented by Bolivia in its Rejoinder. However, we refer the Court to the
review of several of these works by Bolivia’s international experts, DHI (BR,
Vol. 2, Annex 23), who point to the very variable quality of the science that has
been presented.
80
3
1.2 Report background and structure
Through the exchange of written pleadings by Chile (Memorial (CM) and Reply
(CR)) and Bolivia (Counter-Memorial (BCM) and Rejoinder (BR)), substantial
agreement has emerged between the parties and their technical experts concerning
the hydrology of the Silala River system. Key points of agreement were
summarized by us in CR, Vol. 1, p. 101, as follows:
“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.”
Further, it is agreed by both parties that if the channels were to be removed, there
would be a reduction in surface flow, a small increase in wetland evaporation2,
and an increase in cross-border groundwater flow.
An important area of scientific disagreement between the two parties and their
international experts concerns the magnitude of the impacts of the channelization
on trans-border surface and groundwater flows. Bolivia has based its estimates
primarily on modelling by the DHI of the Near Field, a small area of 2.56 square
kilometres (km2) including the Bolivian springs and wetlands and the river
corridor to the international border. Three scenarios were modelled, representing
the status quo, the removal of the channelization, and the removal of
channelization plus the possible long-term growth of peat soils in the wetlands.
2 Both parties agreed that wetland evaporation is a small component of the water balance.
81
4
This modelling produced large estimates of the impact, whereas in the opinion of
Chile’s experts, these effects will be small (CR, Vol. 1, p. 141).
In Bolivia’s Counter-Memorial (BCM, Vol. 1, pp. 52-53), it was stated that if the
channels and drainage works were removed, the Silala River surface flows would
decrease by 30-40%, evaporation from the restored wetlands would increase by
20-30%, and groundwater flow across the border would increase by 7-11%. Chile
noted that the assumed boundary conditions used for the models would exaggerate
the effects, and illustrated the error using a simplified text-book example
calculation. DHI and Bolivia misinterpreted this example calculation (BR, Vol. 1,
p. 33), which was used only to demonstrate the problem with their inappropriate
use of these boundary conditions, and not to produce an estimate of the expected
effect of channelization, but nevertheless DHI accepted that our criticism of the
boundary conditions was justified and its calculations had overestimated the
expected effect (BR, Vol. 5, p. 55). Consequently, DHI presented revised
modelling results in BR, in which the previous results were regarded as an upper
limit: ‘if the channels and drainage mechanisms were removed, cross-border
surface flows in the Silala would decrease by 11% to 33% of current conditions
[…] evapotranspiration from wetlands without canals will increase by 28% to
34% of the reference values, i.e. between 2.8 and 3.4 l/s, while groundwater flows
across the […] border will increase between 4% and 10% as compared to current
conditions’ (BR, Vol. 1, p. 35).
This report addresses continuing and serious concerns for the validity of DHI’s
modelling, summarized in Section 2. The data files used to model the effects of
channelization for Bolivia’s Counter-Memorial were provided by DHI after the
preparation of Chile’s Reply, and in section 3 of the report we analyze these to
show that the modelled results of the different scenarios are not comparable, and
the results that have been presented are seriously misleading to the Court. There
are many issues of concern, but one particularly important factor is that different
82
5
topographies have been used in modelling the different scenarios. The results of
the different scenarios are not comparable, and the differences in topography
between the scenarios are much greater than the dimensions of the channels and
hypothetical peat growth, the impacts of which, the modelling is designed to
quantify. These problems recur with the revised modelling results from Bolivia’s
Rejoinder.
Further, all of DHI’s modelling has been based on Bolivia’s interpretation of the
geology of the Silala River basin, and in section 4 of the report we show that
Bolivia’s understanding of the geology of the Silala River catchment is incorrect,
which invalidates the conceptual hydrogeological models that underlie all of
DHI’s modelling.
In section 5, we address Bolivia’s recent evidence for the impacts of the historical
channelization on the wetlands in Bolivia. While these works, undertaken with the
consent of Bolivia, are the responsibility of Bolivia, it is relevant to the future
management of the Silala River, the restoration of the wetlands, and future
downstream flows, that the reasons for historical changes are clearly understood.
While we do not dispute that there are impacts on wetland ecosystem health to be
expected from channelization, we agree with DHI, who provided an independent
review of some of the supporting documents, that there are several reasons why
the wetlands have changed over time, notably including climatic changes. We
conclude that Bolivia’s documentation includes exaggerated and wholly
unsubstantiated claims about the extent of wetland changes and the role of the
channelization. In section 6 we present our conclusions and answer the questions
posed to us by Chile.
83
6
2 SUMMARY
While there is general agreement between the Parties concerning the hydrology of
the Silala River basin and the nature of potential changes due to historical
channelization in Bolivia, there is important disagreement concerning the likely
magnitude of these effects. Bolivia asserts that there have been large effects of the
channelization on streamflow, while acknowledging that any changes in surface
water flow across the border would be accompanied by associated changes in
groundwater flows across the border. Chile has stated that these estimates are
unrealistic. Given the relatively small reductions in water table depths associated
with the drainage of the wetlands and channelization of the main river, any effects
are likely to be small.
Bolivia’s estimates of change are based on simulations by the Expert consultants,
DHI, which were presented in the Counter-Memorial. Based on the limited
information provided by Bolivia at that time, Chile noted two issues of concern
for the credibility of the modelling. The first concerned the incorrect specification
of boundary conditions for DHI’s Near Field Model (NFM), on which the
estimates of effects were based. The second concerned Bolivia’s incorrect
interpretation of the geology, on which all of DHI’s modelling was based. In
Bolivia’s Rejoinder, DHI accepted the criticism of the boundary conditions and
renamed the previous results as an upper bound estimate of the effects. In parallel,
Chile had requested further information to substantiate Bolivia’s modelling
results, and this was finally provided in February 2019.
In this report we review the additional data and address Bolivia’s revised results
in the Rejoinder. In short, in addition to the concerns noted in Chile’s Reply,
analysis of DHI’s digital data showed a large number of issues of concern, and
many unreported differences between the modelled scenarios, leading to our
conclusion that the results are wholly unreliable, as explained below.
84
7
We also review the new documentation, provided in the Bolivian Rejoinder,
concerning the basis of their geological interpretations and understanding of the
geology of the Silala River basin. These, on detailed examination, were found to
contain significant errors, misinterpretations, to be often confused and with many
internal inconsistencies, which leads us to conclude that the Bolivian modelling is
based on an incorrect understanding of the geology and hydrogeology.
In addition, because it is important for wetland restoration and the prediction of
impacts on downstream flows to understand correctly the impacts of
channelization on the wetlands, we comment on reports by other Bolivian experts,
which provide useful data on historical changes, but make unjustified assertions
concerning the causal effects of the wetlands.
We therefore address the questions posed to us by Chile, below. The first two
questions were addressed to Wheater, and the second two to Peach. The answers
to (i) and (ii) have therefore been drafted by Wheater, and the answers to (iii) and
(iv) by Peach. However, they reflect our joint opinion.
(i) Do the digital data provided by Bolivia in support of the Counter-
Memorial after Chile’s Reply was finalized materially change your assessment of
the modelling by Bolivia’s Experts of the effects of channelization and of the
effects of possible long-term peat growth?
The digital data provided by Bolivia in February 2019 has allowed detailed
analysis of DHI’s modelling, used to support Bolivia’s claims concerning the
impacts of historical channelization of the Silala River in Bolivia, as reported in
Bolivia’s Counter-Memorial. Analysis of model configurations, parameters, input
data and simulation results showed that there were many aspects of the modelling
that gave serious concern for the reliability of the results, in particular for the
modelling of a 2.56 km2 area designated by DHI as the Near Field that includes
85
8
the source water springs and wetlands in Bolivia, and the river channel to the
international border.
The Near Field Model was used by DHI to simulate three scenarios, the Baseline
Scenario (current conditions), a No Canal Scenario (with the channels removed),
and an Undisturbed Scenario (with channels removed and an allowance for the
long term growth of peat in the wetlands), to evaluate the effects of the historical
channelization of the Silala River and its wetlands in Bolivia, and of possible
future peat growth in the wetlands. Bolivia relies on these results to claim large
impacts of historical channelization on Silala River cross-border flows.
We had noted previously, in Chile’s Reply, that there were errors in the geological
interpretation on which the modelling was based (and hence the model
configuration and parameterization), and that there was a major problem in the
choice of boundary condition for the NFM, which caused exaggeration of the
reported effects. However, inspection of the digital data by Chilean hydrologists
(Muñoz et al., 2019) revealed many unreported differences between the models
used for the inter-comparison of scenarios, and in the model boundary conditions
and initial conditions. These unreported differences were compounded by
unexplained methodology, and incorrect assumptions. Perhaps of greatest impact
was the fact that we found that different topographies had been used for the
different scenarios, including different topographies used in the Baseline Scenario
for the modelling of catchment processes (the MIKE-SHE model) and the
modelling of channel flow (the MIKE-11 model). These differences in
topography, of up to 7 m, were far greater than the small changes in channel depth
and peat growth that the models were being used to evaluate, and in themselves
would generate large differences between the scenarios. While the reported model
errors and inaccuracies for the NFM were of a similar magnitude to the effects
being simulated, which in itself casts doubt on the validity of the conclusions from
86
9
the modelling, we conclude that the large effects proposed by Bolivia are mainly
an artefact of these unreported differences between the modelled scenarios.
In short, our assessment of the reliability of the modelling by Bolivia’s Experts
has materially changed. In our professional opinion, the published results of the
effects of channelization and of the effects of possible long-term peat growth are
wholly unreliable and should be disregarded by the Court.
(ii) Do the further modelling and other studies by Bolivia’s Experts, presented
in Bolivia’s Rejoinder, reflect a correct assessment of the magnitude of these
effects?
We were pleased to note from Bolivia’s Rejoinder that DHI accepted our criticism
of the choice of boundary conditions for the NFM, and now considers these
results to set an upper bound for the effects of channelization and possible longterm
peat growth. DHI also provides a lower bound estimate of the simulated
effects, based on a constant flux at the boundary. We agree that this is an
appropriate condition to give a lower bound estimate. However, the simulation
results presented are subject to all of the errors and inconsistencies noted above,
and have errors that are of a similar magnitude to the effects being simulated. In
addition, they differ from the BCM results for the same simulation, which is a
further indication of an unreliable modelling process. Again, in our professional
opinion, these further modelling results are misleading and should be disregarded.
We note that DHI refer to a historical estimate of flow, made in 1922 prior to the
channelization, to support their simulations and conclusions. However, in our
opinion, a single estimate, made at a location that is uncertain, and in a difficult
environment where contemporary measurements have had large errors, cannot be
considered reliable.
87
10
Related studies by other of Bolivia’s experts considered the relationship between
the historical channelization and observed changes in the wetlands. Understanding
the causes of wetland change is important given the desire of both Parties to see
wetland restoration, and to understand the effects on downstream river flows.
While these studies shed important light on some of the changes that have taken
place in the wetlands, they erroneously attribute the single cause of observed
changes to the channelization. Given that the dates of reported changes bear no
relationship to the dates of channelization, it must be concluded that other factors
are playing a significant role. We agree with Bolivia’s experts, DHI, that climate
changes are likely to be one of the more important controls.
(iii) Does the information provided by Bolivia’s Experts, presented in Bolivia’s
Rejoinder, materially change your assessment of the geology and hydrogeology
that underpins Bolivia’s modelling of the effects of channelization and of the
effects of possible long-term peat growth?
Our assessment of the Bolivian interpretation of the geology and hydrogeology of
the Silala River basin has resulted in our clear view that the geological
interpretation that underpins Bolivia’s modelling is substantively flawed. Bolivia
have introduced new geological evidence in the BR that is often confused and
inconsistent. Further, they have ignored Chilean evidence and simply asserted that
Bolivian interpretations are correct.
(iv) What are the implications of this new information, if any, for the validity
of Bolivia’s modelling of the effects of channelization and possible long-term peat
growth?
The geology and hydrogeology of the area around the Bolivian wetland springs
and downstream beneath the Silala River, defined by DHI as the Silala Near Field,
88
11
is complex, and given the limited data, is open to differing interpretations.
However, concerning the geology and hydrogeology, DHI has adopted a set of
models that do not represent the best understanding of reality, and are based on
many erroneous conclusions, assumptions and internally inconsistent
interpretations, in addition to the issues noted above in answer to question (i). The
models are therefore likely to produce results that are grossly in error. We have no
doubt that Bolivia’s estimates of the effects of channelization are highly
exaggerated. This conclusion is in part due to the adoption of boundary conditions
for the NFM that are incorrect, both in their condition and location, as well as the
use of inconsistent and highly variable topography. But it is also due to the
misrepresentation of aquifer geometry, the use of aquifer property distributions
that do not take into account the correct geometry and stratigraphy, and the fact
that the existence of an important shallow aquifer system has been ignored.
The NFM not only does not reflect reality, but is not consistent with the stated
Bolivian conceptual understanding of the groundwater flow regime.
3 MODELLING THE IMPACTS OF CHANNELIZATION ON
THE HYDROLOGY OF THE SILALA RIVER SYSTEM
3.1 Introduction
In Bolivia’s Counter-Memorial, a suite of models was used by DHI to represent
the hydrology of the Silala River and its groundwater catchment area, estimated
by DHI to be 234.2 km2 (Figure 1). Based on DHI’s MIKE-SHE modelling
platform, three models were developed. A model for the Far Field, known as the
Water Balance Model (WBM), was used to simulate the water balance of the
topographic catchment area as well as groundwater flows and residence times
from the extended groundwater catchment. A Near Field Model was developed to
represent the Silala River valley from the international border to the groundwater
89
12
springs, which are the sources of the Silala River surface flows, and the associated
Cajones and Orientales wetlands, an area variously described as 2.7 km2 (BCM,
Vol. 2, p. 301) and as 2.56 km2 (BCM, Vol. 5, p. 13). A third model, the Near
Border Model, was developed to represent the interactions between surface water
and groundwater in more detail, for an area of the Near Field Model between the
confluence of the Orientales and Cajones tributaries and the international border
(BCM, Vol. 5, pp. 72-76). The three model areas are shown in Figure 2. Given
their relevance to the outstanding issues of disagreement between the parties, we
focus on the Far Field and Near Field Models in this report. Results of these
models are summarized in Bolivia’s Counter-Memorial Volume 2 (BCM, Vol. 2,
Annex 17), and the models are presented in more detail in Volumes 3 and 5
(BCM, Vol. 3, Annex 17, Annex E; BCM, Vol. 5, Annex 17, Annexes G and H).
Figure 1. Hydrological catchment and Silala Near Field area defined by DHI
(BCM, Vol. 2, p. 328).
90
Geological faults
Road
lnternallonal border
S1lala canal
c::J Sub catchments
Hydrological catchment
13
Figure 2. Domains covered by the three different DHI models (Muñoz et al., 2019).
With the limited information available from the BCM, we presented a report on
the DHI modelling (Wheater and Peach, 2019) that outlined major concerns for
the validity of the Near Field modelling. Concerning the estimates by DHI that the
natural surface flows without drainage and channelization would be 30-40% less
than the current situation, as stated in BCM (BCM, Vol. 2, p. 266), we noted that
‘In our opinion the very large estimates made by DHI are implausible, and indeed
defy common sense’ (CR, Vol. 1, p. 98).
Given our concerns about the limited information presented in Bolivia’s Counter-
Memorial to define DHI’s models and results, Chile requested access to the digital
files that define the model configuration, parameterization, input and output data,
in a letter to the Agent of Bolivia dated 5 November 2018. A letter dated
91
Caro colk i
od,/ J6n
562
lnacallrl Police
Stati~
CODELCO Intake
Kilometers
M!fcatorProjfction,WGSM
Laguna
Blanco
Laguna
Chila
NEAR BORDER MODEL (NBM)
WATER
BALANCE
MODEL
J&Jon« (WBM)
• 01:ientalrs. - NEAR FIELD MODEL (NFM)
1 ~ BRAZIL
PERU l ~
~ l~
PARAGUAY
14
3 December was received from Mr. Jensen of DHI declining to provide the data.
Dr. Wheater wrote to the Agent of Chile on 19 December, 2018, noting that the
information provided in the Bolivian Counter-Memorial was inadequate to define
the models that were used, or the modelling process that was followed, and
furthermore the detailed results had not been provided.
A further request was sent to the Agent of Bolivia on 21 December 2018, and the
digital data were finally received on 7 February 2019, which was after the
finalization of the editing of Chile’s Reply, that was presented to the Court on
15 February 2019. The data provided essential information to study the
configuration, parameterization and performance of DHI’s models, and a detailed
analysis was made by Chilean hydrologists (Muñoz et al., 2019), under our
supervision. The digital data proved to be extremely illuminating, and show
conclusively that the DHI results are invalid, and should be discounted by the
Court. The full Muñoz et al. (2019) report is attached to Chile’s Additional
Pleading. Below we highlight key findings and conclusions.
3.2 Water balance modelling
The WBM (BCM, Vol. 3, Annex 17, Annex E) was run to provide insights into
the possible groundwater contributing area, the recharge that determines the
magnitude of cross-border surface water and groundwater flows, and groundwater
travel times. Estimates were made of the water balance over the topographic and
groundwater catchments (excluding the Near Field area). The resulting estimate of
recharge was 24 mm/year, though sensitivity analysis indicated a range of
19-49 mm/year3. The mean groundwater travel time was estimated to be
1,500 years, with a range from 50-6,000 years.
3 We note that in DHI’s Provisional Report 3 (Muñoz et al., 2019, Appendix D, p.11), Water
balance of the basin and groundwater aquifer and update of measured flow recharge using the
92
15
The water balance from the WBM was estimated using a simulation period of
17,500 days, or just less than 48 years. Results from one of the files provided in
February 2019, version Silala_model_gw_200m_v12_final.she, are presented in
Table 1. The flows in mm/year and l/s were calculated using the simulation period
and the area of the model’s active cells. The table shows that 198 l/s of
groundwater leaves the model through the NFM boundary.
Table 1. Water balance from the “Water Balance Model” -
Silala_model_gw_200m_v12_final.she version (Muñoz et al., 2019).
It can be noted that even after 48 years, the SHE model has not quite converged to
steady state (there is a net storage change of 170 mm (4 mm/year) over the
simulation period), a point we return to later.
DHI note the need for long-term simulation to estimate the water balance (BCM,
Vol. 3, p. 471). However, it is relevant to point out that while a period of 48 years
smooths the day-to-day and short-term inter-annual variability, this period does
not represent the climate variability over the 6,000 year period that was estimated
by DHI as the maximum groundwater travel time.
Muñoz et al. (2019) also point out that there is an inconsistency between DHI’s
hydrogeological conceptual model and the boundary conditions for the numerical
same model is estimated to be 56 mm/year. No explanation has been provided by DHI for why the
estimate was reduced in their final report.
Cumulative water
depth (mm)
Average depth
rate (mm/year)
Average flow
rate (l/s)
Precipitation -6023 -126 -911
Evapotranspiration 4854 101 734
Recharge (prec.–evap.) -1170 -24 -177
Total storage change -170 -4 -26
Net groundwater boundary
outflow 1309 27 198
Error -30 -1 -5
93
16
WBM (Figure 3). A no-flow boundary is assumed at the model’s south-western
boundary, including a section along the Chile-Bolivia border. This will affect the
modelled groundwater flows to the Near Field, and influence the partition of
recharge between surface water and groundwater.
Figure 3. Boundary conditions of the Water Balance Model (Muñoz et al., 2019).
These issues of failure of the model to reach steady state and of inconsistency in
boundary conditions are of limited importance for the water balance estimation,
but are a recurrent theme for DHI’s modelling, and have much greater
significance for the Near Field modelling, as discussed below.
94
lnacallrl Pollce
Stati~
CODELCO Intake
Kilometers
MN-cmir Pr.ction, WGS M
Laguna
hlca
WATER
BALANCE
MODEL
(WBM)
1.~ BRAZIL
PERU i ~
}
BOLIVIA
~- PARAGUAY
ARGENTINA 2 -
17
3.3 Near Field modelling
3.3.1. The Pleadings to date
DHI’s Near Field modelling lies at the heart of the disagreement between the
Parties. The Near Field Model has been run for three scenarios (BCM, Vol. 5,
pp. 66-72):
i) “Baseline”: represents the current situation with channelization.
ii) “No Canal”: represents the situation without channels.
iii) “Undisturbed”: represents a ‘restored’ situation without channels and with
assumed long-term development of wetland peat soils, of up to 60 cm depth4.
We note that in simulating the Baseline Scenario, DHI’s MIKE-SHE model is
used as the base model to represent the catchment response, including the
unsaturated zone, groundwater, overland flow and evapotranspiration. To
represent flow in the channels, the MIKE-11 surface-water flow model is used.
MIKE-11 links to the MIKE-SHE model and allows exchanges of water between
the channels and the underlying groundwater system. The “No Canal” and
“Undisturbed” scenarios are also based on the MIKE-SHE model, but in these
cases there is no explicit representation of river channels; the MIKE-11 model is
not included.
In Bolivia’s Counter-Memorial, based on the Near Field modelling, DHI stated
that ‘Without canals […] a reduction in surface flow of 30-40% is estimated
compared to current conditions’, ‘The groundwater flow […] at the border
increases by 7-11 %’ and ‘The evapotranspiration increases by 20-30 % by
removing the canals and restoring wetlands. This however, correspond to a
4 We note that Bolivia’s Counter Memorial refers to a “Wetland restoration scenario” (e.g. BCM,
Vol. 5, p. 66) whereas Bolivia’s Rejoinder refers to “Wetland restoration (Undisturbed Scenario)
(BR, Vol. 5, p. 73), and later as simply “Undisturbed” (BR, Vol. 5, p. 79) for the same scenario.
95
18
reduction of only 2-3 l/s in the combined cross border groundwater and surface
water flow.’5 (BCM, Vol. 2, pp. 266-267).
A key aspect of the NFM is that it represents a small area (2.56 km2), located at
the bottom of the groundwater and topographic catchments (Figure 4). Given the
large uncertainties in the recharge to, and properties of, the groundwater
catchment, DHI elected not to run a fully integrated model for the Near Field and
Far Field (BR, Vol. 5, p. 67). The inflow of groundwater to the NFM was
therefore determined by the choice of the boundary conditions for the model. As
shown on Figure 4, a fixed ‘head’, i.e. a fixed groundwater water table elevation,
was specified at the boundaries where groundwater was allowed to enter the Near
Field. Many of the lateral boundaries to the model were defined as no-flow
boundaries, while the lower, south-western boundary had a specified head
gradient as the boundary condition. Specified values of head and head gradient
were based on recent observations.
5 DHI’s italics
96
19
Figure 4. The area of Bolivia’s Near Field Model. The green color represents Fixed
Head boundary conditions, the black color represents No-Flow boundary condition and
the grey color represents Fixed Gradient boundary condition (Muñoz et al., 2019).
In Chile’s Reply (Wheater and Peach, 2019), we noted that the fixed head
boundary condition was inappropriate, because water table conditions at the
model upslope boundary are held constant in the model, while the interior of the
model changes to represent changing channelization and peat growth. In reality
the water table this close to the channel would respond to the changes. The
constant water table boundary condition forces unrealistic changes in the inflow to
the model, which are reflected in unrealistic changes to the outflow. We used
simple calculations for an idealized two-dimensional hillslope cross-section to
show that this boundary condition grossly exaggerates the simulated effect of the
channelization and possible peat growth, perhaps by a factor of 20. We also noted
97
Fixed Gradient
(-0.05)
1200
WATER BALANCE MODEL
(WBM)
Fixed Head
(initial potential head)
Fixed Heacl
(initial potential head)
CHILE
BOLIVIA
Area
Enlarged
20
that there were significant errors in the geology, on which the modelling was
based. Concerning DHI’s proposed changes, we concluded: ‘We agree that these
effects may occur, but find DHI’s large estimates to be implausible. These
estimates are wholly based on hydrological modelling […], which we find to be
fundamentally flawed.’ (CR, Vol. 1, p. 100)
In Bolivia’s Rejoinder (BR, Vol. 5, p. 55), DHI accepted the validity of our
criticism of the boundary conditions: ‘It has been found, however, that the model
boundaries are affected by the changes introduced by the removal of the canals
and that the chosen boundary conditions will therefore also have a bearing on the
produced results for a situation in which the canals have been removed.’ They
now take the position that the sensitivity to the boundary conditions should be
investigated ‘When considering the baseline model and the “no canal”/
“Undisturbed” scenario results, the sensitivity and uncertainty should therefore be
taken into account’ (BR, Vol. 5, p. 55). DHI’s report (BR, Vol. 5, Annex 25)
therefore investigates the sensitivity of the NFM to the assumed upstream and
downstream boundary conditions. The previously adopted fixed head upslope
boundary conditions are now seen as providing an upper bound ‘Assuming that no
changes will occur on the boundary will lead to the largest impacts on the surface
water flows and, hence, such analysis will represent the upper bound’ (BR, Vol. 5,
p. 55). This can be interpreted as acknowledgment that the upper bound results are
unrealistic, and that the impacts will be less than these values. The lower bound
assumption for the upslope boundary condition is that there will be no change in
the groundwater inflow to the NFM, which we agree is a conservative assumption.
In addition, it is acknowledged that the assumed fixed gradient lower boundary
condition will also affect the results.
As a result of the sensitivity analysis, Bolivia’s estimates of the impact of
removing the channels and possible long-term peat growth have been reduced, but
still represent significant effects. The range of results obtained from the scenarios
98
21
without canals from the upper and lower bounds for the upslope boundary
condition is as follows (BR, Vol. 5, p. 56):
• ‘[…] the simulated range of decrease in transborder surface flow […] is
11% - 33%.
• The groundwater flow will increase between 4% and 10% [...]
• The evapotranspiration from the wetlands […] will increase between 28%
and 24% of the baseline values or between 3 and 3.4 l/s.’
The different components of the water balance of the model are specified in Table
2 below.
Table 2. Different components of the water balance of the model (BR, Vol. 5, p. 80).
It will be recalled that the lower bound of DHI’s estimated impact range was
based on no change in the water inflow to the NFM, and from the principle of
conservation of mass, for a steady state simulation, this means no change in total
outflows. However, the results presented by DHI for the lower bound simulation
show a decrease in surface outflow of 16 l/s, an increase in groundwater outflow
of 4 l/s, and an increase in evapotranspiration of 3 l/s. In other words, 9 l/s has
gone missing. This is due to the change in storage and numerical inaccuracy of the
lower bound results, stated by DHI to be 8.4 l/s (see Table 2) , and is more than
half of the supposed impact of the channels the model is intended to estimate.
Further, DHI’s stated error in the Baseline simulation is 11.7 l/s. This means that
99
Table 7-2 Results of the outer bounds of the sensitivity analyses of the upper head boundary conditions as changes
in 1/s from the flow components in the baseline simulation with the canals
Canalised
Changes from canalised conditions (1/s)
situation (1/s)
Baseline Lower Bound Upper Bound
Inflow to model 253.6 -1 -27.9
Surface outflow 149.0 -16 -48.6
Groundwater outflow 106.3 4 10.8
Evapotranspi ration 10.0 3 3.4
Storage and num. inacuracy -11.7 8.4 6.6
22
the combined model errors from the scenarios being compared exceed the
estimated effect of the channels, discrediting the modelling results. Clearly, there
are major issues with the modelling in addition to those arising from the fixed
head upstream boundary conditions that we noted in Chile’s Reply. We turn next
to the files provided by DHI in February 2019, and the analysis of Muñoz et al.
(2019) to provide further insights.
3.3.2. Information from DHI’s model files
The results of any hydrological model depend on assumptions made by the
modellers. These include the detail assumed for the geometry of the system (in
this case including not only the topography but also the representation of channels
and drains and their interactions with surface and subsurface flows), the material
properties used to represent the soils and aquifers and their spatial distribution,
and the assumed boundary conditions of the model, which, as discussed in section
3.3.1, determine the inflows to and discharges from the model. In addition, with a
complex non-linear model, such as the models used by DHI, there are also issues
of numerical performance that are of concern, for example are there instabilities in
the model or other numerical errors that affect the results, and for a steady state
simulation, has the simulation converged to a steady state? As noted above, the
information provided by DHI in Bolivia’s Counter-Memorial was insufficient for
us to review these aspects in any substantive detail. However, the provision of the
associated digital data files by Bolivia in February 2019, and the subsequent
purchase by Chile of licenses to run the models, enabled a comprehensive analysis
to be undertaken by Muñoz et al. (2019).
100
23
3.3.3. Topography of the modelled scenarios
We recall from 3.3.1 above that DHI have modelled the Near Field using three
scenarios:
The “Baseline” Scenario represents the current configuration of the river system,
including the historical wetland drainage channels and main river channelization.
DHI’s MIKE-SHE model is used to represent the catchment response and MIKE-
11 to represent flow in the channels. MIKE-11 links to the MIKE-SHE model and
allows exchanges of water between the channels and the underlying groundwater
system.
The “No Canal” Scenario, to represent the situation with channels removed, and
the “Undisturbed”, or restored wetland, Scenario, to represent the possible long
term development of peat soils (of up to 60 cm depth) are based on the MIKESHE
model alone, the MIKE-11 model is not included.
The lack of a MIKE-11 component for the No Canal and Undisturbed scenarios
means that all surface water flow is routed as overland flow, using the MIKE-SHE
model, using a relatively coarse spatial discretization compared to the MIKE-11
flow routing. This represents a situation in which there would be no surface water
flow channels if the channelization was removed, which is of course incorrect.
Clearly, for the modelled scenarios to be compared, the models must be
comparable. We note from the above scenario descriptions that there are
differences in model configuration (MIKE-SHE plus MIKE-11 for Baseline
versus MIKE-SHE alone for No Canal and Undisturbed). However, one of the
most basic requirements is to have comparable representation of the topography
of the Near Field, while allowing of course for the removal of channels and peat
growth. We were astonished to discover that very different topographies have
been used for the different scenarios, and for the Baseline Scenario a different
topography has been used for the MIKE-11 model and for the underlying
101
24
MIKE-SHE model, with which it links. The differences are illustrated for selected
cross sections in figures 5 and 6. We recall that the aim of the modelling is firstly
to simulate the effects of channels, which in the wetlands are generally less than
0.5 m deep, and in the main channels are all less than 1 m deep (BCM, Vol. 5,
pp. 31-39), and secondly to simulate the effect of possible peat growth, of up to 60
cm. It can be seen from Figure 5 that the topography of the Undisturbed Scenario
(MIKE-SHE) shows differences of almost 7 metres from the Baseline (MIKE-11)
channel bottom in cross section number 3560 and of almost 3 metres from the
Baseline (MIKE-SHE) topography in cross section 3370.
102
25
Figure 5. Ground surface elevations used in the four Bolivian models compared at two
cross sections of the main channel near the international border (Muñoz et al., 2019).
103
4396
4394 _
439J f
~
4290 w
Cross section no. 3560
...
j
2 4 6 12 14 16
4302
4300 ]:
C:
4298 ,Q "~'
Cross section no. 3370
4296 ~w--.:!::==l=t
4294
4292
Dista ce (m)
4 6 8 10 12
- Baseline (MIKE-11) - Undisturbed (MIKE-SHE) • • • No Canal (MIKE-SHE) Baseline (MIKE-SHE)
26
Figure 6. Ground surface elevations used in the four Bolivian models compared at two
cross sections of the main channel in the Orientales wetland. Specifically, in these cross
sections the Baseline and the No Canal topographies from the MIKE-SHE model coincide
and the black dotted line obscures the yellow line (Muñoz et al., 2019).
Clearly, the three scenarios are not comparable, and the large imposed differences
in topography are much greater than the effects that the models are supposed to
distinguish. The No Canal and Undisturbed scenarios are not simply representing
the removal of the channels, but also a large increase in the level of the land
surface on the Bolivian side of the border. The dramatic raising of the ground
surface will have the effect of increasing the groundwater heads at the inflow and
104
BOLIVIA
43-87.8
43-87.6
43-87.4
43-87.2
4387 C
.,0,
~
43,86.8
.,
w
43,86.6
43-86.4
43,86.2
0
Area
Enlarged
Below
Cl
Cross section no. 111 O -
----· --------- -------
2
Distance (m)
-~4-- 6 ----'---8'--~1~0-- 12
4393
4392.8
4392.6 t---~--+----+---
C
4392.4 ..0 , __, __ _, __ -~ ---
4392.21 ~ ---<-, .....
4392
Cross section no. 890
4391.6
4391.4
0
Distance (m)
----' 2---4~- 6 8 10 12 14
- Baseline (MIKE-11) - Undisturbed(MIKE-SHE) • • • NoCanal (MIKE-SHE) Baseline (MIKE-SHE)
27
outflow boundaries, which will in turn reduce the groundwater inflows to the
model, and increase the groundwater outflows, leaving less water available to
appear as surface water flow in their model. In other words, the imposed
topographic changes will give rise to most, if not all, of the simulated effects that
are supposed to be due to the channelization. The results of the scenario
comparisons are clearly meaningless, and should not, in our opinion, be taken
seriously by the Court.
3.3.4. Hydrogeology and boundary conditions
The DHI models depend on DHI’s conceptualization of the hydrogeology of the
Silala River basin, which in turn depends on interpretation of the geology. In
section 4 below, we show that Bolivia’s interpretation of the geology is wrong in
several important respects. For example, errors in Bolivia’s dating of rocks have
led to incorrect specification of the vertical sequence of the rocks. The ignimbrite
deposits, which are the major deep aquifers in the basin, are shown by Bolivia as
underlying the Miocene Volcanics, whereas in fact, the Miocene rocks are older
than the ignimbrites, and cannot therefore underlie the ignimbrites (section 4.2.1;
Muñoz et al., 2019; SERNAGEOMIN, 2019b). One direct consequence of this is
that DHI’s interpretation of the geometry of the aquifer is incorrect. Figure 7
shows a screen shot from one of DHI’s model files, which shows simulated
groundwater flow in part of the Near Field where in reality the low permeability
Miocene deposits will restrict the flow. While the direct effect of these errors in
interpretation of the geology is mis-specification of the geometry of the system,
other effects include consequential errors in the specification of aquifer properties.
Clearly, when the effects to be simulated are small and quite subtle, these basic
errors mean that the simulations will be unreliable.
105
28
Figure 7. Groundwater level maps used in definition of groundwater component
boundary conditions (BCM, Vol. 5, p.19). Black lines represent the piezometric contours,
the polygons filled with grey lines represent the HGU4 unit. The added red arrow
represents the implied groundwater flow through the HGU4 unit (Muñoz et al., 2019).
We turn next to the model boundary conditions. Here we note differences between
DHI’s own interpretation of Bolivia’s groundwater data, and the boundary
conditions used for the NFM. As pointed out in section 4, below, Bolivia’s
groundwater data are open to different interpretations, depending on the
interpretation of the hydrogeology. However, there should at least be consistency
between DHI’s interpretation of the data and their model boundary conditions. In
Figure 8 A) we show DHI’s groundwater contours from Figure 40 (BCM, Vol. 4,
p. 97). We note that groundwater flow direction is determined by the gradient of
groundwater level, and that the direction of flow should be perpendicular to the
contours. We have therefore added arrows to DHI’s figure to show the implied
flow directions. In figures 8 B) and 8 C) we reproduce DHI’s figures 6 and 7
(BCM, Vol. 2, p. 371), which provide a very different interpretation of the
groundwater flow to the springs. However, the boundary conditions used in the
model are quite different from either of these interpretations, as shown in Figure 8
A) (Muñoz et al., 2019). For example, much of the southern boundary of the
model is a no-flow boundary in the model, which is not what is shown by either of
106
(mtta,J
75116800
756e600
751111400
7511&200
7511&000
75115800
75115600
75115AOO
75115200
75115000
&00000 &00500 &01000
lntbal potential head on the saturated zone
&01500 &02000 &02500 &03000 &03500
(meter,
!mewl
Above400
4420 -"30
4410·4'20
4400 - 4410
4390 • . u oo
•380 - 4390
4370-4310
4350-4370
4350-•llO
43'0- 4350
4 3,l0-'3,tO
4320-4330
43 10-4320
4300 - 010
4290-4300
BelowQIO
UndefnedV...,_
29
DHI’s contrasting interpretations of the data. And as noted above, the boundary
conditions for the NFM are also different from those used by DHI for the Water
Balance Model, which was used to simulate flows from the Far Field groundwater
catchment to the Near Field. We conclude that DHI’s model boundary conditions,
which as noted above are crucial to the modelling, are simply not consistent with
the available groundwater data.
107
30
Figure 8. A) Groundwater level contours in the Silala NFM, interpolated from piezometer
wells, spring elevations, and wetlands excavations for soil sampling (Adapted from BCM,
Vol. 4, p. 97). The NFM domain is delimited by the polygon with a black and white
border that shows in black the DHI no-flow boundaries and in white the boundaries
through which water can pass. The blue arrows represent the direction of groundwater
flow interpreted from the contour lines. B) Northern and C) Southern wetlands overall
flow directions. (Adapted from BCM, Vol. 2, p. 371). Note: The text in the lower label of
panel C is: “Drained part of wetland with signs of vegetation changes” (Muñoz et al.,
2019).
108
A Sp,ing
• Piezometers (Groundwater ClevaOon. mHI)
--December 2017 water Lcvets
•-·-· lntemational Border
31
3.3.5. Numerical performance and other modelling issues
In section 3.3.2 above we noted that there are a range of concerns when using
complex, highly non-linear models, related to numerical performance. We recall
(section 3.3.1) that the MIKE-SHE model was used for all scenarios, and that for
the Baseline Scenario, the MIKE-11 model was used to represent the channelized
surface flow. We highlight here some of the more important issues of concern.
A more complete analysis is presented in Muñoz et al. (2019).
Concerning MIKE-SHE, DHI state (BCM, Vol. 5, p. 67) that ‘The integrated
surface water - groundwater model has been set up and run as a steady-state
model.’ This means that there should be no change in the internal storages in the
model, and the model outputs should equal the model inputs. However, all of the
results presented by DHI (e.g. BCM, Vol. 5, p. 67, Table 1; BR, Vol. 5, p. 80,
Tables 7-1 and 7-2) indicate storage change. This means that the simulated water
balances are incorrect. In 3.3.1 we pointed out that the results of the sensitivity
analysis presented in Bolivia’s Rejoinder show that the combination of storage
and numerical inaccuracy accounts for 11.7 l/s in the Baseline Scenario, and
8.4 l/s in the lower bound No Canal simulation, when the lower bound simulation
shows changes in surface flows due to channelization to be 16 l/s. The errors in
the modelling are therefore of the same order as the effects to be modelled.
Clearly the results must be considered unreliable.
A further implication of the failure of the models to reach steady state is that the
initial conditions, which are used to set the initial states and hence storage values
of models, will influence the final results. Muñoz et al. (2019), found that DHI
had used different initial conditions for the different scenarios. Since these
scenarios represent different physical configurations it would be reasonable for
them to have different initial conditions, but two problems arise. Firstly, there is a
methodological inconsistency in DHI’s simulations: the same initial conditions
are used in the Baseline and No Canal scenarios but not in the Undisturbed
109
32
Scenario. Secondly, the differences between the initial groundwater pressures
(represented as an equivalent height of water and known as potential heads) in the
Baseline and No Canal scenarios and the Undisturbed Scenario, shown in Figure 9
vary between -18 m and +16.5 m. These differences only came to light from
inspection of DHI’s digital files, and have not been explained or justified by DHI.
Muñoz et al. (2019) conclude that the very large imposed differences in the initial
conditions mean that the simulations of the three scenarios are ‘neither equivalent
nor comparable’.
Figure 9. Initial potential head difference between the Baseline and No Canal scenarios
and the Undisturbed Scenario. Positive values correspond to locations where the initial
potential head is higher in the Undisturbed Scenario (Muñoz et al., 2019).
Inspection of the data files by Muñoz et al. (2019) also showed unexplained
anomalies in the various model scenario configurations. It is agreed by both
Parties that the springs in Bolivia are fed by groundwater flow, partly from the
topographic catchment and partly from a more extensive groundwater catchment.
A physically based model of the Near Field, the small area around the springs,
should therefore represent the inflow to the springs as a groundwater input across
the NFM boundaries, and this was our understanding of the modelling from DHI’s
reports in BCM and BR. However, from the model files it is apparent that
110
300 600
Mele15
Mffa1o1 PIDjKtion, WGS 84
,_ __ ,._ lnltlal potential head difference: Baseline and No Canal versus Undisturbed
-s-18
Fixed Head
(Initial potential head)
Initial potential head difference (m)
CJ
-12
CJ
-6
CJ
0
CJ
6 -12 -;,18
33
additional water is introduced to the Near Field at the locations of the spring
sources, and this has neither been explained nor justified. 42 l/s of water is
introduced as an external input of “spring recharge” in the Baseline Scenario. In
contrast, 31 l/s is input into the two No Canal scenarios. These amounts appear
arbitrary and are not mentioned in any of the reports available to us, let alone
explained or justified. But clearly a difference of 11 l/s has been introduced into
the scenario comparisons, an amount that is more than half of the simulated
change in surface flows due to the channelization, further enhancing the simulated
effect6.
We turn next to the MIKE-11 model, used in the Baseline Scenario to represent
surface water flow in the channels. There are a large number of issues and
concerns with the model set-up and results, described fully in Muñoz et al. (2019).
We mention a few of the more important issues here.
We noted in 3.3.3 above that a different topography was used for the MIKE-11
and MIKE-SHE models in the Baseline Scenario. Additional topographic issues
were found in the MIKE-11 model for this scenario. It was found that water at
some of the modelled cross sections does not flow through the main channel, as
illustrated in Figure 10. Further, a key parameter in hydraulic models of surface
water flow is the effective roughness of the channel, commonly represented by a
parameter known as Manning’s coefficient (n). As explained by Muñoz et al.
(2019), DHI have used unrealistically large values for this parameter, way beyond
feasible literature values for the types of channel represented here. One result of
this large roughness is that flow velocities are slower than would be expected, and
hence the flow depths are larger, an effect exacerbated in the lowest reaches by an
inappropriate lower boundary condition for the hydraulic model (Muñoz et al.,
6 It is also the case that this injected water is treated differently in the different scenarios. As
explained in Muñoz et al. (2019), in the baseline scenario water is added directly to surface runoff.
In the other scenarios it is added to a MIKE-SHE grid cell as recharge and subsequently
partitioned between overland and subsurface flow, so that the partitioning between surface water
flow and groundwater flow is influenced by this different treatment.
111
34
2019). This perhaps explains why, in DHI’s model, the channel is flooded in
places, and the water flows as over-bank flow (Figure 11). Clearly the
representation of channel flow in the MIKE-11 model does not represent the
reality of channel flow in the Silala River system, and this also affects the
simulation of surface water-groundwater interactions.
112
35
Figure 10. Some of the cross sections where the water does not flow through the main
channel (Muñoz et al. 2019).
113
[rntteirl MAINCANAl 3210 31-03·201615:00:00
003.6
oou
'303.2
.tJOJ.O
002.a
002.6
.tJOl . .t
002.2
.t302.0
oou
"301.6
.tl01.A
'301.2
-14.0 -ll.0 -10.0 ....
[intterJ MAJNCANA.l 3290 3U)3,2016 15:00:00
4299.0
429$.1 Main channel
,nu I 429'.4
42982
4291.0
42'7.I
41.97.6
.t2'7.4
,i,1.2
42'7.0 .... -3.0 ·1.0 -1.0 ... , .. ,.
[mttetl MAINCANAL 3520 31·03·2016 15:00:00
4287.0
428U Main channel
4286.6
418'.4
4U6.2
4286.0 I
421$.8
421S.6
411S.4
428U
428S,O
4284.8
-'llU
4214.4
42842
42&4.0 .... -1.0 ... -3.0 -2.0 -1-0 0.0 1.0 1.0
Water level - - Maximum water level
, ..
...
Main channel I
•..
..• 5.0 ...
, .. •• , ..
Minimum water level
...
lmecerl
36
Figure 11. Cross sections where the channel is flooded (Muñoz et al., 2019).
Returning to the issue of numerical stability, Muñoz et al. (2019) found that the
MIKE-11 simulation results showed abrupt changes in river flow at various
locations, and flow variations along the river that never stabilized, illustrated in
Figure 12. More importantly, there were unexplained differences between the
MIKE-SHE and the MIKE-11 simulated discharges. Overall there was an
unexplained 7 l/s over-estimate of the flows from MIKE-11 compared to MIKESHE
for the current conditions.
114
!In««! MAINCANAL3540 31-03-201615:00:00
•ZKO
•nu
•2n,
......
,12as2
421S0
OIAI
OMI
<12 ...
42142 .,...
"21)1 ... .... ., .. -ao so -40 .u ., 0 ·1.0 00 I 0 ~o 10 •• so •• 70
fffll'terl
lln«•I MAINCAN ... L 3560 31-03-2016 IS:00:00 .,...
O ... l
OIAl
AZIA 1
42'40
OIJ.1
ou,
'2117
421) I,
421U
421)4
•ztJ]
01)2
OaJ1
<12110
.,., '
mu
•1.0 ... -S.O ... • J.o ., . · LO 0.0 LO , .. )0 •• s.o , .. , .. ... lrneterl
-- Water level - - Maximum water level - - Minimum water level
37
Figure 12. Flow variations in two sections of the reach between the Cajones and
Orientales confluence and the international border. (A) Plan view of the NFM domain
that depicts the locations of the two sections analyzed. (B) Discharge time series of the
flow at sections 3550 and 3160. (C) Zoom into the last two days of the flow at sections
3550 and 3160 (Muñoz et al., 2019).
115
JOO 600
Meiers
M«ulDf~n,'MiS84
0.19
0.18
0.17
900
t~1 6 ~
i5015
014 1 I
0.13
02-03-2016
0.170
0.165
0.160
~ 0.155
1
j0,150
g 0.145
0.140
0.135
0.130
30-03-2016
07-03-2016 12-03-2016 17-03-2016 22-03-2016
- MAINCANAL 3550 - MAINCANAL 3160
31-03-2016
- MAINCANAL 3550 MAINCANAL 3160
27-03-2016 01-04-2016
01-04-2016
(
I
I
38
A final inconsistency concerning DHI’s numerical results is that in Bolivia’s
Counter-Memorial, DHI’s estimate for the Undisturbed Scenario (removal of
channels and long-term peat growth) is that the river flow is reduced by 40%
relative to current conditions (BCM, Vol. 2, p. 303). However, in Bolivia’s
Rejoinder, where the same simulations were used to represent the upper bound
estimate of the impacts, DHI reported that the river flow was reduced by 33%
(BR, Vol. 5, p. 56). In attempting to understand this clear contradiction, Muñoz et
al. (2019) found that 10 l/s of additional surface water flow had appeared in the
results for the Undisturbed Scenario presented in Bolivia’s Rejoinder. This was
not explained, but it was included in the Excel spreadsheet named “Water balance
tables – Sensitivity Report.xls”, delivered by DHI along with the files provided to
support the Rejoinder modelling (Muñoz et al., 2019, Appendix C). Once again,
an unexplained and apparently arbitrary change to the model had been made to
change the results.
3.3.6. Flow estimate from Fox (1922)
A measurement of flow in the Silala River was reported by Robert H. Fox, a water
supply engineer, in a paper published in 1922. Fox noted a flow of 11,300 cubic
metres per day, or approximately 131 l/s. This is of course lower than current
flows at the border, and hence DHI (BR, Vol. 5, p. 80) state that this flow
measurement supports their argument that prior to the channelization, flows in the
river would have been lower.
There are however several factors to bear in mind. Firstly, the location of the
measurement is unknown, and as reported by Bolivia and Chile, there are major
changes to the flow rate at different locations along the river, as springs add to the
surface flow, and water is lost from the channel to groundwater. DHI state (BR,
Vol. 5, p. 80): ‘Fox measured 131 l/s […] at a location which from his description
must be pretty close to the present de-siltation chamber.’ However, no
116
39
measurement location was specified in Fox’s paper. A possible interpretation
from the paper is that the measurement was made at the “small dam” (or Intake
N° 1) that was built in 1909 and operative from 1910 onwards, located in Bolivian
territory just below the confluence of the Cajones and Orientales ravines, at
approximately 600 m upstream from the international boundary. This does not
justify Bolivia’s claim that the flow represents the cross-border flow (BR, Vol. 1,
pp. 35-36).
Secondly, the method of measurement is unknown, and hence its accuracy and
reliability. The recent, 21st century, experience from both Bolivia and Chile, is
that the Silala River is a very difficult environment in which to accurately
measure flows. Chile (CM, Vol. 5, p. 247) noted ‘difficulties in flow measurement
due to the extreme conditions of the Silala River’, and Bolivia (BCM, Vol. 2,
p. 395) reported that comparison of the long term flow records from the
permanent gauging stations set up by Bolivia and Chile ‘shows significant
differences in both the mean flow levels and temporal variation. None of the
series from the two sites seems, however, to be free of gauging inconsistencies,
which may be due to the remote locations and harsh climate.’ In addition, shortterm
measurements made by Bolivia in 2017 show ‘inconsistencies both at the
individual gauging points and also when cross-comparing the data.’ (BCM, Vol.
2, p. 395).
It is clear that, even today, the Silala River is a very difficult environment in
which to measure flows, and given that only a single estimate was reported by
Fox, in 1922, using an unknown method at an uncertain location, then little
credibility can be attached to his reported value.
117
40
3.4 Conclusions
The analysis of the digital files provided by Bolivia in February 2019 has
provided extensive and definitive evidence that DHI’s modelling, upon which
Bolivia’s estimates of the effects of historical channelization are based, is fatally
flawed. While there are very many issues of concern, perhaps most striking are
that: 1) completely different topographies were used for the different scenarios,
with the differences in topography very much larger than the relatively small
differences associated with channelization and assumed peat growth, 2) there are
basic errors in the geology and hydrogeology that mean that the geometry and
aquifer properties are wrong, 3) arbitrary amounts of water have been added to the
wetland springs, with different amounts for the different scenarios (and different
partitioning between surface and groundwater), thus adding to the simulated
effects of channelization, and 4) errors and inaccuracies, including changes in
storage, account for similar rates of flow to the effects ascribed to the
channelization and peat growth.
Since our report on the impacts of channelization in Chile’s Reply (Wheater and
Peach, 2019), DHI has provided new results in Bolivia’s Rejoinder. While we
were pleased to note that DHI accepted our criticism of their earlier work for the
BCM, and agreed that their choice of boundary condition biased the results,
nevertheless the revised results from their sensitivity analysis are subject to all of
the errors pointed out above. In addition, the BCM and BR results differ for the
same simulation, and it appears that unexplained amounts of water have been
added to the flow model to produce this change.
In truth, we were shocked by the basic errors made by DHI, and their failure to
report crucially important aspects of the modelling. We conclude that DHIs’
modelling results must be regarded as meaningless, and in our professional
opinion should be discounted by the Court.
118
41
4 ISSUES RELATED TO THE GEOLOGY AND
HYDROGEOLOGY OF THE SILALA BASIN
4.1 Introduction
Bolivia’s estimates of the impact of the channelization of the Bolivian wetlands,
known in Chile as Cajones and Orientales or in Bolivia as Bofedales Norte and
Bofedales Sur, and of the Silala River in Bolivia, have been based on the results
of DHI’s modelling of the surface and groundwater systems of the Silala River,
including its extended groundwater catchment. The numerical models used to
provide these estimates have been constructed from a hydrological and
hydrogeological conceptual model (BCM, Vol. 4, pp. 62-102), which in turn has
relied heavily on the Bolivian understanding of the geology of the area.
In essence, Bolivia’s interpretation of the geology is incorrect in several important
respects, which in turn leads to DHI’s conceptual model of groundwater and
surface water flow being incorrect. In consequence, the construction of DHI’s
numerical models, which are based on that understanding, is also wrong. This will
inevitably result in errors in the estimates of groundwater and surface water flows,
and in particular the results of the NFM, which was used to simulate various
scenarios with and without channelization, as explained in Section 3 above.
A detailed study of the documents listed below which pertain to the geology of the
Silala River groundwater catchment in Bolivia presented to support the BR has
been made by the Chilean geologists (SERNAGEOMIN, 2019b). This is included
in Chile’s Additional Pleading as Appendix 2 to this report.
1. Annex 23.5: F. Urquidi, “Technical analysis of geological, hydrological,
hydrogeological and hydrochemical surveys completed for the Silala water
system”, June 2018. (BR, Vol. 3, pp. 233-332).
2. Annex 23.5, Appendix a: SERGEOMIN (National Service of Geology and
Mining), Study of the Geology, Hydrology, Hydrogeology and Environment
119
42
of the Area of the Silala Springs, June 2000-2001, Final Edition 2003. (BR,
Vol. 3, pp. 333-401).
3. Annex 23.5, Appendix b: SERGEOMIN, “Structural Geological Mapping of
the Area Surrounding the Silala Springs”, September 2017. (BR, Vol. 4, pp. 5-
136).
4. Annex 23.5, Appendix c: Tomás Frías Autonomous University, (TFAU),
“Hydrogeological Characterization of the Silala Springs”, 2018. (BR, Vol. 4,
pp. 137-462).
5. Annex 24: DHI, “Analysis and assessment of Chile’s reply to Bolivia’s
counter-claims on the Silala Case”, March 2019. (BR, Vol. 5, pp. 5-46).
In their report, SERNAGEOMIN, 2019b, list numerous errors and inconsistencies
in the Bolivian documentation, which have a significant impact on the
hydrogeological understanding and have resulted in the flawed construction,
including aquifer parameterization and boundary condition location and type, of
DHI’s NFM. We highlight some of the more important examples of these errors
and inconsistencies in this section, and their consequences, which have led to a
severely flawed representation of the groundwater flow regime and hence
incorrect results of the modelling they rely upon to estimate the effects of
channelization of the Bolivian wetlands and their supporting springs.
In addition to the errors in geological interpretation, DHI’s conceptualization of
groundwater flow has been compared to DHI’s model construction by Muñoz et
al. (2019), and detailed in section 3 of this report. It has also been noted that the
boundary conditions imposed on the NFM are inconsistent with DHI’s own stated
conceptual understanding of the groundwater flow regime. This is a further source
of model error, as has been discussed in section 3.
120
43
4.2 Geology and Hydrogeology of the Silala River wetland springs and
surrounding area
In this section we discuss the errors and internal inconsistencies in the Bolivian
documentation and methodologies used to develop their geological
interpretations, hydrogeological conceptual model and the incorporation of this
understanding into the numerical models prepared by DHI. This discussion is
developed to show the flawed nature of these models and likely consequences for
the accuracy of their estimates of the impacts of channelization.
One of the most basic laws in geology is the “law of superposition” which states
that in undeformed sequences of strata, the oldest strata will be at the bottom of
the sequence. The study of these sequences of rock strata is known as
stratigraphy, which is discussed in section 4.2.1. Another important area for
consideration is the deformation that the rock deposits have undergone over
geological time. This will be dealt with in section 4.2.2 on structural geology. The
branch of science concerned with the origin, structure, and composition of rocks is
called petrology and the detailed descriptions of rocks, their mineral content and
the textural relationships within the rock is known as petrography. There are
several inconsistencies and errors in the description and naming of rock deposits,
which have contributed to the confusion concerning the geological understanding
of the Bolivian geologists (SERNAGEOMIN, 2019b), some of which are briefly
discussed in section 4.2.3.
The Bolivian reports listed above contain important errors and inconsistencies in
the understanding and interpretation of stratigraphy, structural geology and in
petrography. These errors and inconsistencies have led to errors in development of
a hydrogeological conceptual model, upon which DHI relied for the design of
their numerical models, particularly the NFM. DHI’s interpretation of the
hydrogeology is discussed in section 4.2.4 below.
121
44
4.2.1 Stratigraphy
The Bolivian understanding of the stratigraphy of the Silala Basin is significantly
flawed, due in part to errors they have made in the attribution of radiometric dates
to the sequence of ignimbrite deposits (the ignimbrite succession)
(SERNAGEOMIN, 2019b). These form the major deep aquifer of the Silala
groundwater catchment, and hence are fundamental to the modelling of the
groundwater system.
In Bolivia the Ignimbrite succession has been divided into three geological subunits
(Nis-1, Nis-2 and Nis-3, Nis-1 being the lowest sub-unit) of one formation
they call Silala Ignimbrites (Bolivian name, hereafter Bol) (BR, Vol. 4, pp. 43-
51), whereas in Chile only two ignimbrite units have been found. These have been
identified in surface outcrops and borehole cores (SERNAGEOMIN, 2017;
SERNAGEOMIN, 2019a; Arcadis, 2017). However, it is accepted by Chile that
beneath the Bolivian wetland areas there may be, at depth, further ignimbrite
units, but these have not been found in Chile.
Bolivian geologists assign an age range for the ignimbrite succession in Bolivia of
7.8-6.6 Ma (BR, Vol. 3, p. 248; BR, Vol. 4, pp. 39, 43 and 46). A radiometric age
date of 7.8 Ma has been given by Bolivian geologists to the Silala Ignimbrite
(Bol). This age was assigned to a sample from Bofedales Norte (Cajones) and was
said, by Bolivia, to come from a sample collected 16.5 km to the south east
(SERNAGEOMIN, 2019b; Baker and Francis, 1978). However, according to the
original paper of Baker and Frances, 1978, this age should be attributed to
andesitic lavas located 8 km east of the Silala Grande (Bolivian name, Volcán
Apagado in Chile). This age has been extrapolated erroneously to the ignimbrites
of the Silala area. Quite clearly there is considerable confusion demonstrated here
and a significant error has been made by the Bolivian geologists.
122
45
In Chile two units of ignimbrite have been identified; the Silala Ignimbrite
(Chilean name, hereinafter Chi) and the Cabana Ignimbrite (Chilean name,
hereinafter Chi), with ages of 1.61 Ma and 4.12 Ma respectively (CR, Vol. 3,
pp. 202-203). These units have been identified in outcrop in Chile and in a cored
borehole in Chile very close to the international border. Further, the Silala
Ignimbrite (Chi) is seen to cross the border at outcrop and can be traced, using
satellite imagery, into Bolivia. It can be observed on satellite imagery underlying
Pleistocene andesitic lavas (1.48 Ma) (CR, Vol. 3, p. 203) and unconformably
overlying the Silala Ignimbrite Nis-3 (Bol) in Figure 13, which is the uppermost
ignimbrite unit that is described by the Bolivian geologists. They attribute this
ignimbrite unit as having an age of 6.6 Ma (BR, Vol. 4, p. 115). Clearly this is not
the youngest ignimbrite since the Silala Ignimbrite (Chi) unconformably overlies
the Silala Ignimbrite Nis-3 sub-unit (Bol). The presence of two separate
ignimbrite deposits in Chile, one of which can be traced into Bolivia in the area of
the Bofedales Sur (Orientales) has been ignored by Bolivia. Furthermore, the
Silala Ignimbrite Nis-2 sub-unit (Bol), which underlies Nis-3, is described as
Dacitic ignimbrite with andesitic clasts (BR, Vol. 4, pp. 48 and 158). This is
significant because these clasts “[…] pertain to the first Inacaliri volcanic event”
(BR, Vol. 4, p. 158) and this is the reason for the andesitic composition. However,
the Inacaliri volcanic lavas have been dated at 5.84 Ma (Almendras et al., 2002),
which clearly means that the Nis-2 sub-unit must be younger than the Inacaliri
andesitic lavas (5.84 Ma) and therefore the Silala Ignimbrite Nis-3 sub-unit (Bol)
must be even younger since this deposit overlies the Nis-2 sub-unit
(SERNAGEOMIN, 2019b).
123
46
Figure 13. The Silala Ignimbrite (Chi) clearly overlying Silala Ignimbrite Nis-3 (Bol),
located at the East of Bofedales Sur (Orientales). The younger Inacaliri Lavas 2 overlie
the Silala Ignimbrite (Chi) (SERNAGEOMIN, 2019b).
Miocene Volcanics have been identified by Bolivia with an age of 6.04 Ma (BR,
Vol. 4, p. 115) as outcropping in the Silala Chico (Bolivian name, Cerrito de
Silala in Chile) volcanic dome but intruding the ignimbrite succession (Figure 14).
However, this age is older than both the Silala Ignimbrite (Chi) and the Cabana
Ignimbrite (Chi), which clearly must overlie these Miocene Volcanics and cannot
be intruded by them. Also, in the Bolivia’s “Generalized Geological Section of the
Silala Springs” (BR, Vol. 4, p.125), see Figure 15, a further error is found, as the
Silala Chico (Cerrito de Silala) dome unit, with the age 6.04 ± 0.07 Ma, is shown
overlying the Silala Grande (Volcán Apagado) volcano deposits, which have
younger age of 1.74 ± 0.02 Ma. This is clearly impossible. These are just two
examples of the disturbing errors in geological mapping and stratigraphic
interpretation made by Bolivian geologists.
124
\_
Silala lgnimbrite 3 (Bol.)
6.6Ma
Nis-3
Silala lgnimbrite (Chi.)
1.61 Ma
Pliis
300 600 900
Meters
Merutor Pll)jection, WGS84
47
125
/
,_.,.....
A
4400 m...uun.
I.OkU, GEOLOOIC.'Q A ·A'
u-,-.Q,,~
Qclo. Si/ala
~ ~I~
I 11
I
• <
' .
1 --
MAPA No. 1
MAPA. GEOt.OGl4. GEHERAL
0£L MANA.'tTW. Oil. Sil.Al.A
PRO\IINCIA. SUR UPCZ DCL OCPAATWLNTO OE
POTOSI
A'
NII
48
Figure 14. Geological Map of the Silala Study area (BR, Complete Copies of Certain
Annexes, Vol. 2, Annex 23.5 Appendix a, p. 69). (A) Expanded view of the cross section
and (B) legend from Bolivia’s geological map reproduced at Figure 2-1.
126
[!] t l I REFERENCIAS GEOLOGIC/\$
f Q,11" --CMiaOl.. ..-..... .-a.,...,. 1111'.e ilt
C --- u Q,, -- A
p
I, ---- T (!,.1 ---- E C E I
R s ()t ---- .._ .....
N T
E 0 ............. _ A C ~ .. -- R .E
N J N lll -~,-...-....-..,ftOI.
0 0
0 C.
.,__., ..... t,..,. ... .__
,, ..,.. .. t-...............
"""
...,. __
p ... z L
I
0
0 C
E
N
N 0 ,,
E .... , ..... _ "' -- 0
C G .... l u-----u.o.
E -....·.. ~N.-)Ctm!T..ea
N 0
0
0 ... C ~6'111
E
N
0
49
Figure 15. Generalized Geological Section of the Silala Springs (BR, Vol. 4, p. 125)
shows that the ages cited by Bolivia (added for clarity in the figure) do not support the
stratigraphic relationship determined by Bolivian geologists (BR, Vol. 4, p. 115). In the
cross-section, older rocks (Miocene Volcanics) are seen to overlie younger lavas. This is
clearly incorrect. The section also shows the ignimbrites underlying the lavas of the
Silala Chico (Cerrito de Silala) dome, which is also incorrect, since they are younger
than the lavas of the dome. The Chilean age for the Silala Ignimbrite (Chi) is 1.61 Ma
and for the Cabana Ignimbrite (Chi) is 4.12 Ma (SERNAGEOMIN, 2019b)
Further, the three scenarios that were simulated by DHI with the NFM gave
results indicating different inflows for each scenario (BCM, Vol. 5, p. 67, Table 1)
1. Baseline – 253 l/s
2. No Canal – 221 l/s
3. Undisturbed – 216 l/s
127
Valc.ln lnilcilhrt • lm1c:.ilirl Volcomo
EstratovolQn • StriltovOlc.ono
111:nlmbrita Sllala ■ Silaln fgnlmbrite
Manantlal•s del Sll ■la SIL.ala Springs
Domo s11111 Chico • Sllllil Chko Dome
VolC3n Sllala Grande • Silala Grande Volcano
G!NrnAUZtD GWlOGl(Al l!C!ION Of !H[ IIIA!A ll~NGl
Dama Silala Chico
Manantiales
+
+
+
.. i,,, .• ..IIJIIJ
SERGE MIN :,,--=,
OIRLMAR
lgminbrita Silala
50
If the groundwater recharge area remains the same for each scenario, then if the
inflow to the NFM changes as above, the difference between the inflows is lost to
somewhere else. This is highly unlikely to be flowing to Chile as groundwater
through the Miocene Volcanics, and is much more likely to be flowing to Chile in
the Silala and Cabana Ignimbrites (Chi) through the narrow restricted region
mentioned above.
4.2.2 Structural Geology
The DHI conceptual model invokes a fault system, the Silala Fault (BCM, Vol. 4,
pp. 75-76), see Figure 16, which has been included in the NFM as a narrow region
of high hydraulic conductivity, and hence as an important pathway for
groundwater flow in their model. This fault is described as a NNE-SSW trending
structure, which follows the Silala River ravine and crosses the international
border into Chile (BR, Vol. 3, pp. 254 and 283). Analysis of the evidence
presented by Bolivia for this fault has been studied by SERNAGEOMIN, 2019b,
and found wanting. Bolivia contend that this structure controlled the location and
form of the Silala River ravine. DHI write (BR, Vol. 5, p. 24) “Sergeomin (2017)
geological mapping indicates a relatively small displacement of 5 m at the
border”. However a structural analysis (SERNAGEOMIN, 2019b) of this
proposed fault indicates that any displacement would be in the opposite direction
to that indicated by Bolivian geologists. The slightly higher position of the south
eastern ignimbrite layers can be explained more reasonably by the fact that the
Silala Ignimbrites (Bol) unit is tilted to the west as indicated by the Bolivian
geologists (BR, Vol. 4, p. 149). Chile has previously submitted evidence that
demonstrates the fluvial origin of the Silala River ravine (Latorre and Frugone,
2017; SERNAGEOMIN, 2017), but Bolivia appears to persist with an alternative
origin, indicating that the Silala Fault, which Chile has shown does not exist, is
responsible for the location and direction of the Silala River ravine.
128
51
Figure 16. Amended map from DHI, 2018a (BCM, Vol. 4, p. 76, Figure 29) showing in
red the postulated Silala Fault system (CR, Vol. 3, p. 211).
Bolivian geologists indicate that this fault plane, which does not exist, coincides
with the current alignment of the Silala River ravine, and that the vertical and
almost vertical walls of the ravine are “strong evidence of the formation of the
ravines by tectonism and movement of glacial ice and fluvioglacial waters” (BR,
Vol. 3, p. 323). SERNAGEOMIN’s recent studies of the Bolivian documents
(SERNAGEOMIN, 2019b) indicate that even if it existed, which it does not, the
angle of the fault at 48 degrees could not produce the near vertical walls of the
ravine.
129
-- Main canals
- HGU7
0 HGU8
~ HGUl
0 HGU2
□ HGU3
0 HGU4
- HGU5
0 HGU6
BOLIVIA
ArH
Enlar ed
52
Further analysis of Bolivian structural data has revealed the use of inappropriate
data, such as cooling joints, to assess structural trends. Since these are not caused
by structural deformation (tectonic processes) they cannot be used to assess the
structural stress regime. Examination of the principle stress relationships leads to
the conclusion that many of the fractures are a result of compression, which would
mean the fractures would be unlikely to conduct water because they would be
closed (SERNAGEOMIN, 2019b).
Further, Bolivian geologists assert that the ignimbrites of the Bolivian Silala River
springs area have been affected by the “Silala Fault”, which they attribute to an
extension of the Uyuni-Khenayani Fault System (UKFS), in Bolivia (Sempere et
al., 1990; Martínez et al., 1994; Elger et al., 2005). This is a major fault system
found in Bolivia that was active until 10 Ma. Not only does the southerly
extension of this fault system lie 31 km ENE of the Silala River, but it was only
active well before the deposition of any of the ignimbrite deposits found near the
Silala River wetland springs in Bolivia or in Chile, so this tectonic episode cannot
be responsible for fracturing or faulting or any other deformation in those
ignimbrites. The location of the UKFS can be seen on Figure 17.
130
53
Figure 17. UKFS, located 31 km to ENE of the Silala River basin groundwater
catchment. The abbreviation KUFZ in the figure refers to the UKFS, which has been used
in this report (SERNAGEOMIN, 2019b).
4.2.3 Petrography
Numerous inconsistencies and errors have been found in the petrographic
description of the various rock types found in the Silala River basin, in particular
131
AP
studiM! section
CENTRAL ANDES AT 21°S
I\ ,
I ,.
I
I
Salarde
-~ Uyuni
,;!
BOLIVIA
tJ)
<.
San Pablo
de Lipez
- - - - - -·
Ato
ARGENTIN
54
concerning the mis-identification of the mineral assemblages in the three Bolvian
named Ignimbrite sub-units.
For example, a rock sample from the Silala Ignimbrite (Chi) collected in the Silala
River ravine, within 10 m of the border (sample number RSP-52t) is the same
deposit as the Silala Ignimbrite Nis-1 (Bol). This sample collected in Chile, at the
border, is andesitic ignimbrite (SERNAGEOMIN, 2019b). However, in all
descriptions in the Bolivian documents, they refer to Nis-1 as a dacitic ignimbrite.
Another example concerns the comparison of mineral assemblages compiled from
SERGEOMIN (Bolivia), 2017 and TFAU, 2018, in the petrography of the Silala
Ignimbrite Nis-1, Nis-2 and Nis-3 (Bol), which were found to be different
(SERNAGEOMIN, 2019b).
These inconsistences and errors lead to a highly concerning lack of confidence in
the resulting geological interpretations of the Bolivian geologists.
4.2.4 Hydrogeology
The Parties agree that a deep aquifer exists beneath the Silala River groundwater
catchment, although the interpretation of the aquifer geometry is distinctly
different. In Chile the main aquifer is the Cabana Ignimbrite (Chi), with the
younger Silala Ignimbrite (Chi) overlying a Debris Flow deposit that is also
highly permeable (Wheater and Peach, 2017; Peach and Wheater, 2019; Arcadis,
2017; Muñoz et al., 2019). The often highly welded Silala Ignimbrite forms in
part a confining or semi-confining layer.
Evidence has previously been presented (Peach and Wheater, 2019; Arcadis,
2017; SERNAGEOMIN, 2019a; Herrera and Aravena, 2017; Herrera and
Aravena, 2019) to support the existence of a shallower perched aquifer system in
Bolivia and Chile, which has been found in Alluvial deposits in Chile. This has
been acknowledged as a possibility by DHI (BR, Vol. 5, p. 27) and they “clearly
concluded that it is likely that there are two primary and distinct sources of
132
55
groundwater discharging to the Silala springs”. However, they have not included
this understanding in their NFM and state “it has not been the intention to
represent the various sources of groundwater explicitly in the model as it would
not reflect on the split between surface water and groundwater discharge from the
Silala Wetlands” (BR, Vol. 5, p. 27). This is misleading, since groundwater flow
from distinct aquifers would be driven by differing piezometric level distributions.
Since the surface water flows are driven by groundwater/surface water
interactions, which would be head-dependent, they would be likely “to reflect on
the split between surface and groundwater discharge” (BR, Vol. 5, p. 27). No
account has been taken of the two-aquifer system by DHI in their conceptual
model or the construction of their piezometric map (BCM, Vol. 2, p. 293). This
will inevitably lead to an incorrect understanding of the groundwater flow regime,
which would in turn lead to incorrect parameterization of aquifer properties and
assignment of boundary conditions.
Concerning the groundwater flow regime in the NFM area, examination of the
Bolivian conceptual understanding of groundwater flow directions (BCM, Vol. 2,
p. 371, figures 6 and 7) reveals that they are in conflict with those interpreted
from DHI’s potentiometric map, as shown in section 3 above. They are also in
conflict with the location and type of boundary conditions adopted in the NFM,
since flow directions are indicated (Figure 8) across the southern boundary of the
NFM, but this is a no-flow model boundary. This leads to the conclusion that the
NFM does not represent the DHI conceptual understanding of the groundwater
flow regime in the region of the Bolivian wetland springs. If the DHI conceptual
model of groundwater flow directions is incorrect, the estimates of the impact of
channelization made with the NFM are likely to be in error (section 3; Muñoz et
al., 2019).
Review of the Bolivian technical documentation provided in support of the BR
has identified errors of geological mapping, interpretation of the geological
133
56
sequence of deposits, their contact relationships and their petrography. These
errors and inconsistencies mean that the NFM is based to an important degree on
erroneous information concerning the geometry of the deep ignimbrite aquifer and
its parameterization, including the lack of inclusion of a shallower aquifer. The
lateral extent of the ignimbrites, in particular the Silala Ignimbrite (Chi) and
Cabana Ignimbrite (Chi), is constricted by Miocene low permeability volcanics in
its subcrop from the region of the confluence of the Bofedales Norte (Cajones)
and Bofedales Sur (Orientales) ravines down gradient in Bolivia and into Chile,
thus influencing the aquifer geometry and the parameterisation of the NFM. It has
been shown by SERNAGEOMIN (2919b) that the Silala Fault does not exist, thus
a main premise of the DHI conceptual hydrogeological model and hence the NFM
is based on a false interpretation of the geology and its material properties. And
not only is Bolivia’s geology and DHI’s conceptualization incorrect, but in
addition, Muñoz et al. (2019) have shown, using only Bolivian documentation,
that the groundwater flow regime incorporated into NFM is in conflict with DHI’s
own conceptual hydrogeological model.
In summary, it is agreed by both parties that a deep ignimbrite, often very
permeable, aquifer exists in the Silala River groundwater catchment. In Chile this
aquifer is semi-confined or confined, as evidenced by water level monitoring and
drilling (Arcadis, 2017). However, it is not agreed that the ignimbrite succession
lies beneath the Miocene Volcanics, as proven in the discussion above and in
detail in SERNAGEOMIN (2019b), thus the areal extent of the ignimbrite aquifer
is constrained by the Miocene Volcanics, making the aquifer geometry quite
different from that proposed by Bolivia. It is also not agreed that the so-called
Silala Fault exists at all and it cannot be invoked to support a narrow zone of high
permeability along the Silala River ravine. And it is not agreed that the shallow
aquifer system can be ignored in understanding the groundwater flow to the
Bolivian springs.
134
57
4.3 Conclusions
A large number of errors and inconsistencies have been found in Bolivia’s
geological mapping and structural analysis. Consequently, DHI’s interpretation of
the hydrogeology and its implementation in the NFM contains many errors, the
most important of which are listed below:
1. An error in the assignment of a radiometric date to establish the age range of
the Silala Ignimbrites (Bol) leading to an incorrect interpretation of the
stratigraphy, which has important impacts on aquifer geometry and the
distribution of permeability in the NFM, the ignimbrite aquifer having a much
more restricted areal extent than proposed by Bolivia
2. The stratigraphic position and contact relationships between different deposits
has been ignored leading to an incorrect formulation of aquifer geometry and
distribution of aquifer properties,
3. Bolivia has ignored the existence of the Silala Ignimbrite (Chi) and Cabana
Ignimbrite (Chi) in their establishment of the Ignimbrite stratigraphy. The
Silala Ignimbrite (Chi) is highly welded, and outcrops unconformably over
much older Ignimbrites in the Bofedales Sur (Orientales). The Cabana
Ignimbrite (Chi) is highly permeable. Both have a limited lateral extent and
are constrained between two hills of Miocene low permeability volcanics, thus
limiting the flow of groundwater through this region. This would impact on
the NFM parameterization and aquifer geometry incorporated into the NFM
4. The Silala Fault, invoked as a high-permeability groundwater pathway by
DHI, does not exist and could not be related to tectonic events that took place
millions of years before the ignimbrites or the Miocene Volcanics were
deposited and cannot be used to specify and narrow high-permeability zone
running down the Bofedales Norte (Cajones), Bofedales Sur (Orientales) and
Silala River ravine.
135
58
5. The Bolivian structural analysis is flawed leading to erroneous interpretations
in the structural geology, and presence and location of open fractures able to
conduct groundwater, so there is a likelihood of incorrect assignment of
aquifer properties in both conceptual and numerical modelling
6. The Bolivian descriptions, and naming, of rock types was found to be
confused, inconsistent and often in error, leading to the conclusion that the
geological mapping in Bolivia is flawed, which means that the
hydrogeological understanding is likely to be flawed
7. DHI has ignored Chilean evidence of a shallow aquifer system and its own
acknowledgement of two sources of groundwater to the Bolivian wetland
springs, leading to incorrect interpretation of the groundwater water table
distribution (potentiometry)
8. The Bolivian conceptual model and potentiometric contours used for the NFM
are in conflict and represent different interpretations of the groundwater flow
regime
9. The NFM boundary conditions, with respect to both type and location, are in
conflict with the Bolivian hydrogeological conceptual model and the
potentiometric contours used for the NFM, which would lead to a different
flow regime than that simulated in the NFM.
All of these listed issues affect the representation of groundwater/surface water
interaction in the NFM and in turn affect the estimation of the impact of the
channelization on surface and groundwater flows.
This list is disturbing and leads to the conclusion that the modelling which has
been used to support and justify the Bolivian estimates of the impact of
channelization on the surface and groundwater flows from the Bolivian wetlands
at headwaters of the Silala River is incorrect. The models developed by DHI as
136
59
Bolivia’s expert advisors are based on an incorrect understanding of the geology
and hydrogeology of the Silala River surface and groundwater catchments.
5 ERRONEOUS STATEMENTS AND FLAWED ATTRIBUTION
CONCERNING WETLAND DEGRADATION
Chile acknowledges that drainage of the wetlands in Bolivia may have had an
impact on their spatial extent and associated ecosystem health (e.g. CR, Vol. 1,
p. 128). However, while the impacts of wetland drainage in Bolivia, undertaken
with the consent of Bolivia, are the responsibility of Bolivia, it is relevant to the
future management of the Silala River, the restoration of the wetlands, and future
downstream flows, that historical changes are clearly understood. Hence, we wish
to point out to the Court that the Bolivian Rejoinder exaggerates the historical
decline in the area of the wetlands and imputes a causal relationship between
changes in wetland extent and the effects of channelization that is unsupported by
scientific data.
Firstly, we consider the wetland area. In Bolivia’s Rejoinder estimates of
historical and current wetland area are reiterated from the Counter-Memorial
(BCM, Vol. 5, p. 163), which were based on the Ramsar report: ‘prior to the
channelization, the Silala River region within Bolivia was covered by high
altitude wetlands known as bofedales that spanned an estimated 141,200 m2 (or
14.1 Ha). Today, those wetlands have shrunk to a mere 6,000m2 (or 0.6 Ha).’
(BR, Vol. 1, p. 18). We note some confusion on the part of Bolivia with respect to
the area affected - their footnote 58 in the Rejoinder (BR, Vol.1, p. 18) that
supports this comment also cites the FUNDECO report which states that the area
affected is 11.48 Ha (BR, Vol. 3, p. 61). More importantly, the estimate of 0.6 Ha
for the current wetland area is clearly erroneous and disagrees with Bolivia’s own
experts’ assessment. The area of active wetland is subject to large seasonal and
137
60
inter-annual variability, as clearly stated by Bolivia’s experts, Torrez Soria et al.
(BCM, Vol. 3, p. 73) ‘bofedals have a seasonal behavior, which can be evidenced
by presenting dramatic changes according to the season and the annual
variations’. They cite Bolivia’s expert Castel (2017), who reports an analysis of
Landsat data from 1975 to 2000. This shows high variability of active wetland
area, varying between 2 and 8 Ha, with a slightly increasing trend. Chile’s original
analysis (CM, Vol. 4, p. 37), also based on Landsat data, similarly showed strong
seasonal variability, with a possible increasing trend, although with somewhat
larger estimates of wetland area. Higher resolution (10 m) satellite data, presented
in Chile’s Reply (CR, Vol. 1, pp. 127-139) enabled more accurate assessment of
the active wetland area to be made, albeit over a limited period (July-November
2018). This confirmed the strong seasonal variability in vegetation activity (the
property sensed using the satellite NDVI data) and showed the maximum extent
of active wetland area for the two Bolivian wetlands of 9.9 Ha, with the lowest
active area recorded (3 Ha) at the beginning of that period. We also reiterate that
these data, presented in the Reply, show that active wetland vegetation currently
occupies the entirety of the valley bottoms.
We reiterate that Bolivia’s repeated estimate of the current wetland area (0.6 Ha)
is not credible, and note that DHI agrees with us. ‘It seems that the areas in the
Ramsar report are not reflecting the full wetland’ (BR, Vol. 5, p. 41). We
therefore note that Bolivia’s statement (BR, Vol. 1, p. 47) that ‘the reduction of
the bofedal area as a consequence was of approximately of 94%’ is grossly
inaccurate with respect to the areas involved. We note in passing that Bolivia (BR,
Vol. 1, p. 46) attacks Chile’s analysis of the high resolution data summarized in
CR, Vol. 1, p. 136, ‘The growth from 2.86 Ha to 7.50 Ha in a period of two
months is an overestimation that reveals flawed calculation that cannot be
reasonable accepted.’ Bolivia apparently ignores the evidence of its own experts,
Torrez Soria et al. (BCM, Vol. 3, p. 73) and Castel (2017), and confounds the
138
61
normal expansion and contraction of areas of active wetland vegetation that takes
place as the seasons progress, with the growth of wholly new areas of vegetation.
Secondly, we address the issue of causality, and consider the statement (BR,
Vol. 1, p. 50) that ‘The scientific evidence shows that the hydraulic works
generated the fragmentation of the bofedals’. The FUNDECO studies, on which
this statement is based, are themselves extremely confused. One strand of
evidence summarized by FUNDECO (BR, Vol. 3, p. 142) is based on
geochemical survey. They state that the geochemical survey shows gradual
desiccation. ‘This period of desiccation began around 1908, which is a clear sign
of the effects that canalization had on the Silala springs.’ But the date of
installation of drainage works in the wetlands is 1928. So these changes were
initiated 20 years before the channelization. A similar statement is made by
FUNDECO, based this time on pollen analysis, ‘From 1908 onwards, a gradual
desiccation process took place and is evidenced by the change in palynomorph
composition […]’ (BR, Vol. 3, p. 142), so again they note a change that predates
the drainage construction. They also note that ‘this desiccation process reached its
climax around 1950 […]’ (BR, Vol. 3, p. 142), which to our knowledge does not
coincide with any channel changes, and again points to other causes of change
than channelization. A further strand of evidence comes from analysis of soil
profiles (BR, Vol. 3, p. 155):
“The first change of soil stratum happened near the point with dating
that corresponds to an age of 680 – 862 years ago, and therefore in a
period prior to the canalization process […] This type of stratum
continued until the next change, which occurred between 1960 –
1980 years (Figure 6). This period corresponds to the critical period,
due to a lack of water, that occurred approximately 50 years after the
canalization works were installed—which remain in operation until
the present time.”
So multiple strands of evidence point to major changes in the wetlands at points in
time that bear no relationship to the channelization works. As pointed out by DHI,
139
62
in their review of this and other Bolivian science reports (BR, Vol. 2, p. 99), ‘the
study does not look further into other climatic aspects such as potential climatic
changes during the last 100-120 years to assess whether such changes could have
been the cause for some of the observed changes to the bofedales.’
We agree with DHI (BR, Vol. 2, p. 78) who state, with respect to the changes in
flow paths associated with the channelization of the wetlands, that ‘this change
may7 be one of the main reasons for the alterations in the habitats, which have
taken place during the last century.’ But in our opinion, and as also suggested by
DHI, climate changes are likely to have been a major factor, to which we add that
the impact of human activities may also be relevant (FUNDECO report (BR,
Vol. 3, p. 38), from interviews with members of local communities that the
bodefales were used for cattle grazing). At any rate, what is clear is that causal
association with channelization of the wetlands has not been proved and
statements to that effect by Bolivia and its experts are simply untrue.
6 CONCLUSIONS
(i) Do the digital data provided by Bolivia in support of the Counter-
Memorial after Chile’s Reply was finalized materially change your assessment of
the modelling by Bolivia’s Experts of the effects of channelization and of the
effects of possible long-term peat growth?
The digital data provided by Bolivia in February 2019 has allowed detailed
analysis of DHI’s modelling, used to support Bolivia’s claims concerning the
impacts of historical channelization of the Silala River in Bolivia, as reported in
Bolivia’s Counter-Memorial. Analysis of model configurations, parameters, input
data and simulation results showed that there were many aspects of the modelling
that gave serious concern for the reliability of the results, in particular for the
7Our italics
140
63
modelling of a 2.56 km2 area designated by DHI as the Near Field that includes
the source water springs and wetlands in Bolivia, and the river channel to the
international border.
The Near Field Model was used by DHI to simulate three scenarios, the Baseline
Scenario (current conditions), a No Canal Scenario (with the channels removed),
and a Undisturbed Scenario (with channels removed and an allowance for the long
term growth of peat in the wetlands), to evaluate the effects of the historical
channelization of the Silala River and its wetlands in Bolivia, and of possible
future peat growth in the wetlands. Bolivia relies on these results to claim large
impacts of historical channelization on Silala River cross-border flows.
We had noted previously, in Chile’s Reply, that there were errors in the geological
interpretation on which the modelling was based (and hence the model
configuration and parameterization), and that there was a major problem in the
choice of boundary condition for the NFM, which caused exaggeration of the
reported effects. However, inspection of the digital data by Chilean hydrologists
(Muñoz et al., 2019) revealed many unreported differences between the models
used for the inter-comparison of scenarios, and in the model boundary conditions
and initial conditions. These unreported differences were compounded by
unexplained methodology, and incorrect assumptions. Perhaps of greatest impact
was the fact that we found that different topographies had been used for the
different scenarios, including different topographies used in the Baseline Scenario
for the modelling of catchment processes (the MIKE-SHE model) and the
modelling of channel flow (the MIKE-11 model). These differences in
topography, of up to 7 m, were far greater than the small changes in channel depth
and peat growth that the models were being used to evaluate, and in themselves
would generate large differences between the scenarios. While the reported model
errors and inaccuracies for the NFM were of a similar magnitude to the effects
being simulated, which in itself casts doubt on the validity of the conclusions from
141
64
the modelling, we conclude that the large effects proposed by Bolivia are mainly
an artefact of these unreported differences between the modelled scenarios.
In short, our assessment of the reliability of the modelling by Bolivia’s Experts
has materially changed. In our professional opinion, the published results of the
effects of channelization and of the effects of possible long-term peat growth are
wholly unreliable and should be disregarded by the Court.
(ii) Do the further modelling and other studies by Bolivia’s Experts, presented
in Bolivia’s Rejoinder, reflect a correct assessment of the magnitude of these
effects?
We were pleased to note from Bolivia’s Rejoinder that DHI accepted our criticism
of the choice of boundary conditions for the NFM, and now considers these
results to set an upper bound for the effects of channelization and possible longterm
peat growth. DHI also provides a lower bound estimate of the simulated
effects, based on a constant flux at the boundary. We agree that this is an
appropriate condition to give a lower bound estimate. However, the simulation
results presented are subject to all of the errors and inconsistencies noted above,
and have errors that are of a similar magnitude to the effects being simulated. In
addition, they differ from the BCM results for the same simulation, which is a
further indication of an unreliable modelling process. Again, in our professional
opinion, these further modelling results are misleading and should be disregarded.
We note that DHI refer to a historical estimate of flow, made in 1922 prior to the
channelization, to support their simulations and conclusions. However, in our
opinion, a single estimate, made at a location that is uncertain, and in a difficult
environment where contemporary measurements have had large errors, cannot be
considered reliable.
142
65
Related studies by other of Bolivia’s experts considered the relationship between
the historical channelization and observed changes in the wetlands. Understanding
the causes of wetland change is important given the desire of both Parties to see
wetland restoration, and to understand the effects on downstream river flows.
While these studies shed important light on some of the changes that have taken
place in the wetlands, they erroneously attribute the single cause of observed
changes to the channelization. Given that the dates of reported changes bear no
relationship to the dates of channelization, it must be concluded that other factors
are playing a significant role. We agree with Bolivia’s experts, DHI, that climate
changes are likely to be one of the more important controls.
(iii) Does the information provided by Bolivia’s Experts, presented in Bolivia’s
Rejoinder, materially change your assessment of the geology and hydrogeology
that underpins Bolivia’s modelling of the effects of channelization and of the
effects of possible long-term peat growth?
Our assessment of the Bolivian interpretation of the geology and hydrogeology of
the Silala River basin has resulted in our clear view that the geological
interpretation that underpins Bolivia’s modelling is substantively flawed. Bolivia
has introduced new geological evidence in the BR that is often confused and
inconsistent. Further, they have ignored Chilean evidence and simply asserted that
Bolivian interpretations are correct.
(iv) What are the implications of this new information, if any, for the validity
of Bolivia’s modelling of the effects of channelization and possible long-term peat
growth?
The geology and hydrogeology of the area around the Bolivian wetland springs
and downstream beneath the Silala River, defined by DHI as the Silala Near Field,
143
66
is complex, and given the limited data, is open to differing interpretations.
However, concerning the geology and hydrogeology, DHI has adopted a set of
models that do not represent the best understanding of reality, and are based on
many erroneous conclusions, assumptions and internally inconsistent
interpretations, in addition to the issues noted above in answer to question (i). The
models are therefore likely to produce results that are grossly in error. We have no
doubt that Bolivia’s estimates of the effects of channelization are highly
exaggerated. This conclusion is in part due to the adoption of boundary conditions
for the NFM that are incorrect, both in their condition and location, as well as the
use of inconsistent and highly variable topography. But it is also due to the
misrepresentation of aquifer geometry, the use of aquifer property distributions
that do not take into account the correct geometry and stratigraphy, and the fact
that the existence of an important shallow aquifer system has been ignored.
The NFM not only does not reflect reality, but is not consistent with the stated
Bolivian conceptual understanding of the groundwater flow regime.
144
67
7 REFERENCES
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.
Arcadis, 2017. Detailed hydrogeological study of the Silala River. (Chile’s
Memorial, Vol. 4, Annex II).
Baker, M.C.W. and Francis, P.W., 1978. Upper Cenozoic volcanism in the central
Andes - ages and volumes. Earth and Planetary Science Letters, 41 (2), 175–187.
Castel, A.P., 2017. 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. Technical Analysis and Independent
Validation Opinion of Supplementary Technical studies concerning the Silala
Springs. (Bolivia’s Rejoinder, Vol. 2, Annex 23).
Danish Hydraulic Institute (DHI), 2019. Analysis and assessment of Chile’s reply
to Bolivia’s counter-claims on the Silala Case. (Bolivia’s Rejoinder, Vol. 5,
Annex 24).
Danish Hydraulic Institute (DHI), 2019. Updating of the mathematical
hydrological model scenarios of the Silala spring waters with: Sensitivity analysis
of the model boundaries. (Bolivia’s Rejoinder, Vol. 5, Annex 25).
Elger, K., 2003. Analysis of deformation and tectonic history of the Southern
Altiplano Plateau (Bolivia) and their importance for plateau formation, Phd
Thesis. Scientific Technical Report STR; 03/05, Potsdam: Deutsches
GeoForschungsZentrum GFZ.
145
68
Elger, K., Oncken, O. and Glodny, J., 2005. Plateau-style accumulation of
deformation: Southern Altiplano. Tectonics, 24 (4), TC4020.
Fox, R.H.,1922. The Waterworks Department of the Antofagasta (Chili) &
Bolivia Railway Company. South African Journal of Science, Volume XIX, 120-
131. (Chile’s Memorial, Vol. 3, Annex 75).
Fundación para el Desarrollo de la ecología (FUNDECO), 2018. Study of
Evaluation of Environmental Impacts in the Silala. (Bolivia’s Rejoinder, Vol. 3,
Annex 23.3).
Fundación para el Desarrollo de la ecología (FUNDECO), 2018. Study of
Evaluation of Environmental Impacts in the Silala, Palynology. (Bolivia’s
Rejoinder, Vol. 3, Annex 23.4).
Herrera, C. and Aravena, R., 2017, Chemical and Isotopic Characterization of
Surface Water and Groundwater of the Silala Transboundary Basin, Second
Region, Chile”. (Chile’s Memorial, Vol. 4, Annex III).
Herrera, C. and Aravena, R., 2019. 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).
Latorre, C. and Frugone, M., 2017. Holocene Sedimentary History of the Río
Silala (Antofagasta Region, Chile). (Chile’s Memorial, Vol. 5, Annex IV).
Martínez, C., Soria, E., Uribe, H., Escobar, A. and Hinojosa, A., 1994. Estructura
y evolución del altiplano suroccidental: el sistema de cabalgamientos de Uyuni-
Khenayani y su relación con la sedimentación terciaria. Revista Técnica de
Yacimientos Petrolíferos Fiscales Bolivianos, 15 (3-4), 245-264.
Muñoz, J.F., Suárez, F., Sanzana, P. and Taylor, A., 2019. Assessment of the
Silala River Basin Hydrological Models Developed by DHI. (Chile’s Additional
Pleading, Vol. 2, Annex XV).
146
69
Peach, D.W. and Wheater, H.S., 2019. Concerning the Geology, Hydrogeology
and Hydrochemistry of the Silala River Basin. (Chile’s Reply, Vol. 1).
Ramsar Convention Secretariat, 2018. Report Ramsar Advisory Mission N°84,
Ramsar Site Los Lípez, Bolivia. (Bolivia’s Counter-Memorial, Vol. 5, Annex
18).
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).
Sempere, T., Herail, G., Oller, J. and Bonhomme, M., 1990. Late Oligocene-early
Miocene major tectonic crisis and related basins in Bolivia. Geology, 18, 946-949.
SERGEOMIN (Bolivia), 2003. Study of the Geology, Hydrology, Hydrogeology
and Environment of the Area of the Silala Springs. (Bolivia’s Rejoinder, Vol. 3,
Annex 23.5, Appendix a).
SERGEOMIN (Bolivia), 2017. Structural Geological Mapping of the Area
Surrounding the Silala Springs. (Bolivia’s Rejoinder, Vol. 4, Annex 23.5,
Appendix b).
SERNAGEOMIN (Chile), 2017. Geology of the Silala River Basin. (Chile’s
Memorial, Vol. 5, Annex VIII).
SERNAGEOMIN (Chile), 2019a. Geology of the Silala River Basin: An Updated
Interpretation. (Chile’s Reply, Vol. 3, Annex XIV).
SERNAGEOMIN (Chile), 2019b. A Brief Review of the Geology Presented in
Annexes of the Rejoinder of the Plurinational State of Bolivia. (Chile’s
Additional Pleading, Vol. 2, Annex XVI).
Tomás Frías Autonomous University (TFAU), 2018. Hydrogeological
Characterization of the Silala Springs. (Bolivia’s Rejoinder, Vol. 4, Annex 23.5,
Appendix c).
147
70
Urquidi, F., 2018. Technical analysis of geological, hydrological, hydrogeological
and hydrochemical surveys completed for the Silala water system. (Bolivia’s
Rejoinder, Vol. 3, Annex 23.5).
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., & Peach, D. W., 2019. Impacts of Channelization of the Silala
River in Bolivia on the Hydrology of the Silala River Basin (Chilean Reply,
Vol. 1).
148
149
Statement of Independence and Truth
1. The opm1ons I have expressed in my Report 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
Expert Report that accompany the Additional Pleading of the Republic of Chile and in
giving oral evidence, if required to give such evidence. r have complied and wilf
continue to comply with that duty.
3. I have done my best, in preparing the Report, 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 Report. l consider that all the matters on which l have
expressed an opinion are within my field of expertise.
4. ln preparing this Report, l 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, l
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. Howard Wheater
Hydrological Engineer
03 September 2019
150
Statement of Independence and Truth
1. The opm10ns I have expressed in my Report 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
Expert Report that accompany the Additional Pleading 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 Report, 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 Report. I consider that all the matters on which I have
expressed an opinion are within my field of expertise.
4. In preparing this Report, 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. Denis Peach
Hydrogeologist
03 September 2019
VOLUME 2
LIST OF ANNEXES
ANNEX Nº TITLE PAGE Nº
ANNEX TO THE ADDITIONAL PLEADING
Annex 100 100.1 Note from the Agent of the Republic of Chile
to the Agent of the Plurinational State of
Bolivia, 27 May 2019
(Original in English)
3
100.2 Note from the Agent of the Plurinational State
of Bolivia to the Agent of the Republic of
Chile, 17 June 2019
(Original in English)
5
ANNEXES TO THE EXPERT REPORT
Annex XV Muñoz, J.F., Suárez, F., Sanzana, P. and
Taylor, A., 2019. Assessment of the Silala River
Basin Hydrological Models Developed by DHI
7
Annex XVI SERNAGEOMIN (National Geology and
Mining Service), 2019. A Brief Review of the
Geology presented in Annexes of the Rejoinder
of the Plurinational State of Bolivia
187
151i

CERTIFICATION
I certify that the annexes and reports filed with this Additional Pleading are
true copies of the documents referred to.
Ximena Fuentes T.
Agent of the Republic of Chile
16 September 2019
153