Review of Hydrol ogical Investigations Carried Out Under the … · 2015-03-06 · Review of Hydrol ogical Investigations Carried Out Under the Menindee Lakes Memorandum of Understanding

Review of Hydrological Investigations Carried Out Under the Menindee Lakes Memorandum of Understanding

Final ReportJanuary 2012

NSW OFFICE OF WATER

Looking west to the Main Weir, Lake Wetherell (centre, right) and Lake Pamamaroo (centre, top). Under some of the water savings options being considered, the water storage function of these two Lakes would be retained whilst Lakes Menindee and/or Cawndilla (located further south) would be decommissioned or only occasionally used for water storage. Such proposals will significantly alter the current water management regime of the system including reducing evaporation which averages in excess of 400GL annually from all four Lakes.

NSW OFFICE OF WATER

Review of Hydrological Investigations Carried Out Under the Menindee Lakes Memorandum of Understanding

Final Report

January 2012

Prepared by: Bewsher Consulting Pty Ltd 6/28 Langston Place, Epping NSW 2121 Australia P O Box 352, Epping NSW 1710 Australia Telephone: (02) 9868 1966. Facsimile: (02) 9868 5759 Web: www.bewsher.com.au E-mail: [email protected] 003137068. ABN 24 312 540 210

The concepts and information contained in this document are the property of Bewsher Consulting Pty Ltd. Use or copying of this document in whole or in part without the written permission of Bewsher Consulting Pty Ltd constitutes an infringement of copyright. This report has been prepared on behalf of and for the exclusive use of Bewsher Consulting Pty Ltd’s client. Bewsher Consulting Pty Ltd accepts no liability or responsibility whatsoever in respect of any use of or reliance upon this report by any third party.

Review of Menindee Lakes MOU Investigations Bewsher Consulting Pty Ltd Final Report, January 2012 J1952R_4

TABLE OF CONTENTS Page

1. INTRODUCTION 31.1 BACKGROUND 3

1.1.1 Water Savings at Menindee Lakes 3 1.1.2 Menindee Lakes Memorandum of Understanding (MOU) 4 1.1.3 Hydrological Investigations under the MOU 4 1.1.4 Objective of this Review 5

1.2 PREVIOUS INVESTIGATIONS into water savings 5 1.2.1 Menindee Lakes Ecological Sustainable Development Project 5 1.2.2 Darling River Water Saving Project 6

1.3 TERMS OF REFERENCE (TOR) 6 1.4 ACTIVITIES UNDERTAKEN DURING THIS REVIEW 10

1.4.1 Documents Reviewed 10 1.4.2 Interviews and Discussions During the Review 11

2. DISCUSSION OF ISSUES 122.1 THE WATER SAVING OPTION CONSIDERED IN THIS REVIEW 12

2.1.1 Water Saving Option Recommended by CSIRO at end of 2010 12 2.1.2 Process Leading to Recommendation of this Option by CSIRO 13

2.2 MEANING OF NO "IMPACT" REQUIREMENTS OF MOU 14 2.2.1 MOU Requirements 14 2.2.2 Some Impacts are Inevitable 14

2.3 HYDROLOGICAL MODELLING FRAMEWORK USED TO ASSESS OPTIONS 15

2.4 NUMERICAL CRITERIA TO ASSESS "NO IMPACTS" 15 2.5 ENVIRONMENTAL FILLING OF LAKES MENINDEE AND

CAWNDILLA 16 2.6 CSIRO DOCUMENTATION PROVIDED FOR REVIEW 17 2.7 SECURING BROKEN HILL'S WATER SUPPLY 19 2.8 CAPACITY TO WITHSTAND DROUGHTS WORSE THAN HISTORICAL 19 2.9 FLOODING IMPACTS 20 2.10 IMPACTS ACKNOWLEDGED IN CSIRO'S DOCUMENTATION 21 2.11 OTHER IMPACTS OR POTENTIAL IMPACTS OF CONCERN

TO NOW 22 2.12 PROPOSED USE OF SAVED WATER AND ITS PROPOSED

LICENCE 23

3. REVIEW FINDINGS 253.1 TOR 1: Adequacy of the modelling assumptions 25 3.2 TOR 2: Review conclusions from the CSIRO reports 25 3.3. TOR 3: Report on the impacts and risks 26 3.4 TOR 4: Comment on the effectiveness of any mitigation measures 27 3.5 TOR 5: Review the risk associated with the preferred option modelled 27 3.6 TOR 6: Comment on issues relating to the new entitlement proposed 28 3.7 TOR 7: Examine the interactions with Broken Hill's supply arrangements 28 3.8 TOR 8: Discuss alternate options 29

Review of Menindee Lakes MOU Investigations Bewsher Consulting Pty Ltd Final Report, January 2012 J1952R_4

TABLE OF CONTENTS (Cont'd) Page

FIGURES

FIGURE 1 — The Menindee Lakes System in Western NSW 3

FIGURE 2 — Lake Menindee Storage Frequency using data sourced from the CSIRO 'No Impacts' Report 18

FIGURE 3 — Lake Cawndilla Storage Frequency using data sourced from the CSIRO 'No Impacts' Report 18

TABLES

TABLE 1 — Part B Schemes 7

APPENDICES

APPENDIX A — NOW's Criteria For 'No Impacts'

APPENDIX B — Storage Frequency Plots

Review of Menindee Lakes MOU Investigations 3 Bewsher Consulting Pty Ltd Final Report, January 2012 J1952R_4

1. INTRODUCTION

1.1 BACKGROUND

1.1.1 Water Savings at Menindee Lakes

The Menindee Lakes water storage is the largest on the Darling River with a capacity of 1731 Gigalitres (GL), average depth of 7m and a surface area of 460km2. The storage consists of nine interconnected lakes including four large natural Lakes: Wetherell, Pamamaroo, Menindee and Cawndilla (see Figure 1). They supply water for Broken Hill and downstream users.

Figure 1: The Menindee Lakes System in Western NSW. (Source: CSIRO)

The Menindee Lakes system was constructed in the 1960’s to: provide secure water supply for Broken Hill; provide water for irrigation and farm supplies in the lower Darling River; meet stock and domestic water requirements along the Great Darling Anabranch,

and supplement the River Murray System, including the supply of water to South

Australia.

The Menindee Lakes storage is owned by the NSW Government and leased to the Murray-Darling Basin Authority (MDBA) as a part of the operation of the River Murray system. Under the Murray-Darling Basin Agreement, when the volume stored in the Menindee Lakes rises to 640 GL, the water is managed by the MDBA to supply water to southern NSW, Victoria and South Australia. When the volume falls below 480 GL, the water is managed as a drought reserve by NSW. This arrangement is known as the ‘640/480 rule’.


Because of the shallow 'saucer' shaped characteristics of the Lakes, their large surface area and their location within a hot arid region, over 400GL of water is lost by evaporation on average every year. For over two decades, numerous investigations have been carried out to in order to identify practical, cost-effective and environmentally responsible means of reducing evaporation. These have involved a mix of structural options (i.e. changes to existing infrastructure) and non-structural options (i.e. changes to water management operations).

1.1.2 Menindee Lakes Memorandum of Understanding (MOU)

In July 2010, the Australian and NSW Governments entered into a Memorandum of Understanding (MOU) for the cooperative investigation and subsequent implementation of key water reform initiatives in NSW, including improved security of Broken Hill’s water supply and changes to the Menindee Lakes operational arrangements. This agreement followed on from the completion of the Darling River Water Savings Project Part B which is described briefly in Section 1.2.2 below.

Under the MOU, the two Governments agreed to work together to confirm the infrastructure and operational changes at Menindee Lakes which will achieve the greatest reduction in evaporation losses to secure water savings for environmental flows while protecting the local environment and heritage. The MOU also stated that:

"a precursor to achieving savings was the provision of an alternate, secure water supply for Broken Hill and the demonstration that neither the existing entitlement holder's water security nor the environment will be adversely impacted as result of the changed operations at Menindee Lakes"1

Of key importance to NSW was that the MOU required that:

"any operational changes at Menindee Lakes must be shown to have no directly attributable adverse impact on the water security of existing water entitlement holders at Menindee Lakes and the Lower Darling River, or the Murray River"2

1.1.3 Hydrological Investigations under the MOU

Hydrological investigations of the water savings options were undertaken by CSIRO under engagement from the Commonwealth Department of Sustainability, Environment, Water, Population and Communities (SEWPaC). The majority of the hydrological modelling work itself was carried out for CSIRO by the Murray Darling Basin Authority (MDBA). CSIRO's work focussed on the more extreme options (B1- B3) identified in Part B of the Darling River Water Savings Project and which involved very dramatic changes to the hydrological regime of Lakes Menindee and Cawndilla. These options had previously been identified to have the higher benefit-cost ratios and higher water savings.

The investigation also adopted an environmental filling regime for Lakes Menindee and Cawndilla that was designed to promote pre-development environmental values in these Lakes rather than maintaining the existing environment and amenity.

CSIRO's work was undertaken by Mr Geoff Podger and culminated in three reports (hereinafter referred to as the 'CSIRO reports'):-

1 MOU, Paragraph 2 2 MOU, Paragraph 8


Darling Water Savings: Options for Environmental Filling, Nov 2010. Darling Water Savings: Options for Environmental Filling – Supplementary Report 1,

Nov 2010. Darling Water Savings: Options for Environmental Filling – No Impacts, Dec 2010.

These CSIRO reports suggest that the option(s) proposed by SEWPaC will have little or no impacts and are consistent with the requirements of the MOU.

The reviewer understands that after examining the CSIRO reports at the end of 2010 the NSW Government had concerns that the proposed changes would have 'directly attributable adverse impacts' which they considered were contrary to the requirements of the MOU.

1.1.4 Objective of this Review

Given these concerns, late in 2010 the NSW Office of Water (NOW) prepared a brief for an independent review of the hydrologic investigations to determine whether the proposals were in accordance with the MOU. A key objective of the review was to determine whether there were 'no directly attributable impacts'.

Bewsher Consulting Pty Ltd was subsequently commissioned in the second quarter of 2011 to carry out the review. The Terms of Reference of the review are listed in Section 1.3below.

It is important to understand that the review was to be undertaken based on the information available at December 2010. The reviewer understands that whilst further hydrologic investigations have been undertaken subsequently by CSIRO for SEWPaC during 2011, these were not known to NOW when the Terms of Reference for this review were drafted and consequently, these investigations are not the subject of this review.

1.2 PREVIOUS INVESTIGATIONS INTO WATER SAVINGS

Since 1995, the NSW Government has undertaken a series of investigations to identify potential structural works to improve the efficiency of the Menindee Lakes. The most significant of these investigations are outlined below.

1.2.1 Menindee Lakes Ecological Sustainable Development Project

In 2002, the Menindee Lakes Ecological Sustainable Development Project identified that relatively minor efficiency improvements, of approximately 10,000ML per year could be achieved through structural works costing about $30M. These included:

improving the outlet capacity of Lake Menindee to the Darling River;

installation of a small block bank and regulator between Lake Menindee and Lake Cawndilla to retain small and medium inflows in Lake Menindee; and

pumping the residual pool of Lake Menindee to Lake Pamamaroo.

An Environmental Impact Statement was completed for these works in 2005. However, no further progress has been made.


1.2.2 Darling River Water Saving Project

Part A of the Darling River Water Saving Project was completed in 2007 and identified, but did not test, a handful of water saving options for the Menindee Lakes system.

Part B of the Darling River Water Saving Project commenced in 2008 and built on the work completed in 2009. The Part B report presented six (6) options for further consideration by Government.

The hydrologic modelling adopted an environmental filling pattern that would require either Lake Menindee and/or Lake Cawndilla to be filled periodically once in every 5 to 7 years on average.3

The six (6) options represented potential alternatives at Menindee Lakes, ranging from:

Option B1 (the most extreme), which could achieve in excess of 200GL/yr of average savings, but would involve lowering the threshold for NSW control of the Lakes, and keeping Lakes Menindee and Cawndilla effectively dry; to

Option B6 (the least extreme), which would achieve more modest evaporation savings of about 34GL/yr on average, would keep the current threshold for NSW control of the Lakes, but would keep Lake Cawndilla effectively dry except for periodic environmental filling.

Impacts on the water availability to users in the Murray Valley under a more extreme dry sequence than historical were not assessed.

The recommended Part B options are summarised in Table 1.

1.3 TERMS OF REFERENCE (TOR)

NOW asked for this review to address the following eight terms of reference (TOR) based on the information available at the end of December 2010. Further NOW advised the reviewer that the criteria listed in the paper prepared by Andrew Brown and Lindsay White4 were agreed as a basis for assessment of impacts associated with the water savings project:

TOR 1: Report on the adequacy of the modelling assumptions

In particular assumptions and model settings used to: maximise evaporation savings; maximise entitlement that might be created from the evaporation savings; and minimise downstream impacts.

TOR 2: Review conclusions from the CSIRO reports

In particular whether: all the assessment criteria have been addressed; the criteria of the MOU have been achieved; and the data would support any alternative conclusions.

3 Note that the environmental filling pattern used in the Part B study differs from the various filling patterns used by CSIRO. See also Footnote 10. 4 This paper is listed below in Section 1.4.1 (g) and reproduced in Appendix A.


Table 1 � Part B Schemes

Description B1 Scheme 1

Do not fill Lake Menindee and Cawndilla, including:- Pamamaroo drainage channel Rapid drawdown to 150GL Alternative supply for Broken Hill

Water Savings Estimated Savings 248GL

Operational Risk Time < 100GL 18%

B2 Scheme 2Environmental fill of Menindee and Cawndilla, including:- Pamamaroo drainage channel Rapid drawdown to 200GL Alternative supply for Broken Hill



B3 Scheme 3(Scheme 2 + Enlarged Outlets) Environmental fill of Menindee and Cawndilla, including:- Alternative supply for Broken Hill Pamamaroo drainage channel Increased outlet capacity of

Menindee/Cawndilla (optional) Rapid drawdown to 200GL

Water Savings Estimated Savings1 125GL


(Location shown diagrammatically)


Table 1 � Part B Schemes (Continued) Description

B4 Scheme 4Environmental fill of Cawndilla,including:- Menindee/Cawndilla bank &

regulator. Pamamaroo drainage channel Rapid drawdown to 200GL Alternative supply for Broken Hill



B5 Scheme 5(Scheme 4 + Enlarged Outlets) Environmental fill of Cawndilla, including:- Menindee/Cawndilla bank and

regulator, Pamamaroo drainage channel, Increased outlet capacity of

Menindee/Cawndilla (optional) Rapid drawdown to 200GL Alternative supply for Broken Hill


Operational Risk Time < 100GL 7% (Location shown diagrammatically)

B6 Scheme 6Environmental fill of Cawndilla, including:- Menindee/Cawndilla bank and

regulator, Pamamaroo drainage channel, increased outlet capacity of

Menindee/Cawndilla (opt) No change to NSW drawdown Alternative supply for Broken Hill


Operational Risk Time < 100GL 2% (Location shown diagrammatically)


TOR 3: Report on the impacts and risks of the proposed options modelled on:

NSW water users; Broken Hill; the Darling Anabranch; the Lake Wetherell floodplain; and Victoria and South Australia.

TOR 4: Comment on the effectiveness of any mitigation measures proposed in the CSIRO reports

TOR 5: Review the risk associated with the preferred option modelled

In particular, with respect to prolonged dry sequences and the ability to maintain supply to NSW Murray and Lower Darling irrigators.

TOR 6: Comment on issues relating to the form and quantum of the new entitlement proposed to be created

In particular, how the future use of that entitlement may impact on different user groups depending on where the environmental water licence is utilised.

TOR 7: Examine the interactions of the proposed option with current and future Broken Hill water supply arrangements

In particular: how an MAR scheme might impact on Lake operations and hydrology; how BH could source additional entitlement to allow for aquifer losses and how this

might impact on Cap issues; and whether a surface storage solution with backup from groundwater might impact on

Lake operations and how this might impact on potential savings.

TOR 8: Discuss alternate options

In particular whether Options B4 – B6 would provide better outcomes in terms of user and environmental impacts while still achieving significant water savings.


1.4 ACTIVITIES UNDERTAKEN DURING THIS REVIEW

1.4.1 Documents Reviewed

At the commencement of the review a number of documents were provided by NOW for consideration during the review. These comprised:

(a) Assessing Environmental Water Requirements. Chapter 8 Lower − Darling River System. MDBA. July 2010;

(b) Memorandum of Understanding between the NSW Government and the Commonwealth of Australia. July 2010. This MOU was for "the cooperative investigation and subsequent implementation of key water reform initiatives in NSW, including Broken Hill's urban water supply and Menindee Lakes operational arrangements";

(c) HMWG responses to paper tabled by Mr Steve Raft (NOW) and Mr Andrew Brown (NOW), Tuesday, 31 August 2010;

(d) Environmental Watering Requirements of Lakes Menindee and Cawndilla − Working Document. Mark Southwell and Ian Burns, MDBA Environmental Planning Branch. September 2011;

(e) Menindee Lakes Storages − Ian Burns Environmental Filling Diagram. Date unknown;

(f) No Adverse Impact on the Environment. Notes prepared by NOW. Date unknown;

(g) Measuring No Downstream Impacts and Lake Failure. Andrew Brown & Lindsay White. Version 5. 15 September 2010. (This document is reproduced in Appendix A);

(h) Menindee Lakes Hydrological Modelling − Towards A Specification (16 September 2010). Menindee Lakes Project Hydrological Modelling Working Group. Version 7;

(i) Darling Water Savings: Options for Environmental Filling. CSIRO: Water for a Healthy Country National Research Flagship. November 2010. Written by Geoff Podger for SEWPaC;

(j) Darling Water Savings: Options for Environmental Filling, Supplementary Report 1.CSIRO: Water for a Healthy Country National Research Flagship. November 2010. Written by Geoff Podger for SEWPaC;

(k) Darling Water Savings: Options for Environmental Filling, No Impacts. CSIRO: Water for a Healthy Country National Research Flagship. December 2010. Written by Geoff Podger for SEWPaC; and

(l) Menindee MOU − Stochastic Reporting. Andrew Davidson and Andrew Brown. NSW Office of the Water. December 2010. Version 2.

The three CSIRO reports listed in (i), (j) and (k) above are subsequently referred to as the 'CSIRO reports' and were prepared over the period November−December 2010. The latter two reports, i.e. (j) and (k) include refinements and supplements to the preceding report. The final CSIRO report prepared in December 2010, i.e. (k), contains the key final version of the modelling which is of prime interest to this review. This report is referred to as the 'No Impacts' report below.


1.4.2 Interviews and Discussions During the Review

In addition to the review of the above documents, the reviewer conducted interviews with the following personnel:

(a) Mr Geoff Podger, CSIRO, Canberra;

(b) Mr Andy Close, MDBA, Canberra (by telephone);

(c) Mr Andrew Brown, NSW Office of Water, Queanbeyan; and

(d) Messrs Chris Ribbon and Steven Raft, NSW Office of Water, Parramatta.

The assistance of the above personnel in carrying out this review is gratefully acknowledged. All of the contacts with these individuals including the personal interviews were conducted in a frank and open manner and the reviewer has no reason to believe that any important information was withheld.


2. DISCUSSION OF ISSUES

2.1 THE WATER SAVING OPTION CONSIDERED IN THIS REVIEW

2.1.1 Water Saving Option Recommended by CSIRO at End of 2010

There have been many dozens of water saving options investigated in the past. The past investigative work has allowed the range of potential options to be narrowed. It is not the purpose of this review to examine these investigations. Rather it focuses on the option presented in the 'No Impacts' report referred to in item (k) of Section 1.4.1.

The reviewer understands this was the recommended option at the end of 2010 and it is this option together with the hydrologic modelling work surrounding it that is the subject of the current review5.

Of all the options listed in Table 1, CSIRO's recommended option most closely resembles Options B1−B3, but also includes further modifications. The key components of the recommended option comprise:

(a) environmental filling rules for Lakes Menindee and Cawndilla in accordance the proposed Basin Plan as it was at the end of 2010. These rules are discussed further in Section 2.5 below and involve these two Lakes being kept empty or at much lower storage volumes, compared with current operations;

(b) a 185/185 NSW/MDBA operating rule6;

(c) the Lake Menindee outlet regulator capacity increased to 14,400ML/d;

(d) improvements to the Lake Pamamaroo outlet channel;

(e) provision of a managed aquifer recharge (MAR) scheme to provide security of supply to Broken Hill once the proposed changes to Menindee Lakes were implemented; and

(f) revised operational target curves for the filling and emptying of Lakes Pamamaroo and Wetherell.

The CSIRO have advised that this option achieves water savings of up to 174GL/year7. The licence associated with these savings is 125.6GL/y (LTCE)8.

5 The reviewer also understands based on advice from CSIRO that further refinements and modifications to this option have been made during 2011. The option discussed in this review is therefore not the latest and may not be the 'best' option currently recommended by CSIRO. Nevertheless as this review focuses on the recommended option and its associated modelling work that were current the end of 2010, it is not appropriate for the review to consider the 2011 option(s). 6 This would be a modification to the existing 640/480 NSW/MDBA operational rule associated with the Menindee Lakes pursuant to Clause 99 of the Murray-Darling Basin Agreement. Under the existing rule, operational control transfers from the MDBA to NSW at a volume of 480 GL when the Lakes are emptying, and transfers back to the MDBA at a volume of 640 GL when the Lakes are filling. 7 The manner in which the savings quantity has been calculated has not been specified in the CSIRO reports. It is likely to be a combination of the increased flow at Weir 32 and the increased flow at the Lake Cawndilla outlet regulator. 8 Long term cap equivalent (LTCE). The licence would be a supplementary type licence allowing for extraction at Weir 32 at a maximum rate of 25,000 ML/d when Lake Victoria is spilling and Weir 32 spills are above 1000ML/d with a maximum annual limit of 347GL. This is equivalent to a 101GL/y (LTCE) licence at the South Australian Border.


Further CSIRO have advised, during discussions with the reviewer, that this option was chosen ahead of other options (such as the alternatives in Table 1) primarily because of the much superior benefit-cost ratio and higher volumetric saving.

2.1.2 Process Leading to Recommendation of this Option by CSIRO

The following is an abbreviated summary of the key steps leading to selection of CSIRO's recommended option. (This information was provided during the reviewer's discussions with CSIRO)9:

(a) The outcome of the Darling Water Savings project and subsequent cost-benefit analysis was the recommendation of six water saving scenarios. SEWPaC selected the two options that had a benefit-to-cost ratio greater than one (i.e. Options B1 and B2 – refer Table 1).

(b) CSIRO reviewed the two recommended options using the most up-to-date models and found:

(i) closing off Lakes Menindee and Cawndilla and changing the ‘640/480 rule’ to a ‘150/100 rule’ (i.e. Option 1) resulted in water savings of approximately 248GL/yr – water savings partially consumed by downstream users; and

(ii) implementing the proposed environmental refilling rules for Lakes Menindee and Cawndilla and changing the ‘640/480 rule’ to a ‘210/200 rule’ (‘Option 2’), saved 125GL/yr.

(c) CSIRO recommended that these options be re-analysed to consider proposed limits on diversions and to determine the impact of lower NSW operating rule thresholds and climate change scenarios on downstream users. Based on this initial work, CSIRO found that a combination of 10GL of General Security and 290GL of Supplementary Access licence and a ‘180/180 rule’ achieve the desired objectives.

(d) Following on from this initial work, the Hydrological Modelling Working Group (HMWG) agreed on a set of indicators for the modelling – to represent lake flooding and ‘no impact’ – and recommended that CSIRO extend their analysis to explore the likelihood of the Lakes being unable to deliver water.

(e) CSIRO investigated a range of environmental filling scenarios and included revised MDBA environmental filling rules in the models. The scenarios were compared against the agreed indicators and the impacts in drought and flood years were also evaluated.

(f) From the new modelling undertaken using the revised environmental filling rules10, CSIRO found:

(i) a ‘150/150 rule’ combined with an increase in the Menindee outlet capacity to 14,000 ML/day and securing of Broken Hill’s water from groundwater sources, resulted in 165GL/year of water savings and minor downstream impacts. (This is reported by CSIRO in their second November 2010 report). Water savings were

9 NOW subsequently advised the reviewer that the work described in (a), (b) and (c) was undertaken by CSIRO for SEWPaC without NOW's involvement. Further NOW advised that the iterations in (e) and (f) were largely as a result of NOW raising concerns with CSIRO about the resultant impacts. 10 NOW advised that following the preparation of the environmental filling rules used during the Part B project, the MDBA initially revised these with the objective of maintaining the current environment of the Lakes. Subsequently they undertook further revisions to the rules in conjunction with the CSIRO modelling work.


partially consumed by downstream users and the remainder flowed to South Australia;

(ii) by changing to a ‘185/185 rule’ there were slightly higher water savings of 175GL/year and minor impacts to NSW Murray allocations (-1% in November that recovers by June) and average Morgan salinities (1.1EC which could be offset by Commonwealth salinity credits). (This was reported in CSIRO's 'No Impacts' report in December 2010); and

(iii) the largest impact discovered in this analysis was a 120 GL lowering of the maximum restriction for SA entitlements in the most extreme low-inflow scenario. This came about because of a period of low inflows in the Darling Basin coinciding with a similar period of low inflows in the Murray System. CSIRO considered the probability of such an occurrence to be extremely low.

2.2 NO IMPACT REQUIREMENTS OF MOU

2.2.1 MOU Requirements

The MOU places various requirements on the parties to ensure the water saving option(s) do not cause impacts of various types and magnitudes, i.e.:

(a) Clause 2: − "A precursor to achieving� savings is� demonstration that neither the existing entitlement holders' water security nor the environment will be adversely impacted� ";

(b) Clause 8: − "any operational changes that Menindee Lakes must be shown to have no directly attributable adverse impact on the water security of existing water entitlement holders of Menindee Lakes and the Lower Darling River, or the Murray River".

Further in assessing the options there is a requirement for hydrological modelling to be undertaken which properly considers the implications of changes in operations, as noted in the following clause:

(c) Clause 7: − "� work will include further hydrological modelling to assess any implications of changes to the operation of Menindee Lakes, including downstream impacts and implications for the Murray River system management";

2.2.2 Some Impacts are Inevitable

In considering the impacts requirements of the MOU it must be realised that in a complex water management system such as that within Murray-Darling Basin it is impossible to alter infrastructure or management rules without impacting something.

Further some impacts, and particularly environmental impacts, can be both positive and negative and judgement may be required to determine whether there is a net beneficial or non-beneficial effect. For example, changes to the filling frequency of Lake Cawndilla and Menindee to produce a more natural behaviour may likely have some negative impacts on vegetation that has established over the last five decades and is dependent on the current


filling/emptying regime. The MDBA also consider proposed regime changes will likely have beneficial effects as a new vegetation regime becomes established.11

The issue of the significance and acceptability of any adverse impacts is briefly discussed further in Section 2.4.

2.3 HYDROLOGICAL MODELLING FRAMEWORK USED TO ASSESS OPTIONS

The simulation of hydrological changes to the basin as a result of implementation of the proposed options has been carried out using the MSM-BigMod modelling framework.

This modelling framework is well established and has been utilised for all major hydrologic assessments in the Basin for over a decade. The models have also been subject to various expert reviews and were utilised during the CSIRO's Sustainable Yields Project (which itself was also subjected to peer reviews).

Discussions with both CSIRO and NOW during the course of the review have not identified any concerns with the modelling framework. Consequently the adequacy of the modelling framework has not been considered further during the current review.

2.4 NUMERICAL CRITERIA TO ASSESS 'NO IMPACTS'

Following on from the reviewer's discussions with NOW and CSIRO, the fact that some impacts will result from implementation of the proposed water savings option does not seem to be disputed (nor does the modelling tool used to quantify the impacts appear to be in dispute).

However there is a dispute about what numerical criteria were agreed to be ‘acceptable’ by the project’s Hydrological Modelling Working Group (HMWG) in about September 2010.

There appear to be two view’s about the agreed numerical criteria:

(a) Andrew Brown & Lindsay White’s discussion paper “Measuring No Downstream Impacts and Lake Failure" Version 5 – 15/9/2010”; or

(b) CSIRO's list reproduced usually as Table 1-1 in their three reports in November and December 2010. Other graphical information was also provided in Appendix A of the CSIRO reports.

Resolution of this dispute on impact criteria is not a hydrological issue but rather has to do with meeting procedures and what was actually agreed at the HMWG and in other discussions between the parties.

11 This review is focusing on hydrological issues and it is beyond its terms of reference to identify whether in cases such as Lakes Cawndilla and Menindee, the vegetation changes produced by the recommended water savings option will mean the "environment will be adversely impacted" (Clause 2 of MOU). Nevertheless clearly it is the view of Mark Southwell and Ian Burns (refer document in Section 1.4.1(d) above) that the proposed changes to Lakes Cawndilla and Menindee present a unique opportunity to restore a large area of highly productive wetland habitat in the Basin. Their views however were made in the absence of consideration of any social and economic impacts associated with the proposed changes to the lakes.


For the purposes of this review however Brown and White's list in (a) has been adopted consistent with advice from NOW and the Terms of Reference of this review.

It is noted that the CSIRO's list in (b) is a largely a subset of Brown and White's list in (a) and some of the 'aspirational' criteria in (a) have not been included in (b). These excluded requirements (or requirements that were addressed in insufficient detail) comprise:

performance in achieving the environmental watering of Lakes Menindee and Cawndilla;

full consideration of whether a loss of drought reserve in the Lakes to service Broken Hill's requirements, would impact on the water security of local water users during a repeat of the worse historical drought or a more severe event;

consideration of the influence of the proposed managed aquifer recharge (MAR) scheme on local water security;

the manner in which the saved water held by the Commonwealth Environmental Water Holder (CEWH) is to be used and its impact on water management systems within the Lakes, the Lower Darling and the Anabranch, and the Murray;

ability to meet the Weir 32 target flows consistent with the Water Sharing Plan;

changes in the lengths of dry periods including recovery times;

ability to deviate from lake storage targets particularly in regard to the annual drawdown of Lake Wetherell from its floodplain to conserve its wetland vegetation12;

consideration of the range of potential impacts on Lower Darling, NSW Murray, Victorian Murray and SA water users, beyond those evident from the nominated indicators;

consideration of changes in the periods when the Lakes are under NSW control; and

changes in flow distributions, losses and salinities at key locations within the Lower Darling and the Murray.

NOW advised the reviewer that interpreting model results against a 'no impacts' test, was always recognised to be problematic, especially when dealing with the effected communities. Further they had consistently advised CSIRO that it would be preferable if the option(s) were adjusted to achieve small positive impacts.

2.5 ENVIRONMENTAL FILLING OF LAKES MENINDEE AND CAWNDILLA

The rationale behind the proposed environmental watering of Lakes Menindee and Cawndilla is discussed in the document prepared by Southwell and Burns and referred to in Section 1.4.1(d) above. The proposals include for the retention of 'current operation rules' in Lakes Wetherell and Pamamaroo, and the re-instatement of a more variable flow regime for Lakes Menindee and Cawndilla. The flow regime for Lakes Menindee and Cawndilla is

12 This is understood to be current operational practice and is not simulated in MSM-BigMod. NOW has concerns that the operational flexibility to achieve this drawdown will be reduced under the changed operating regime of the Lakes.


specified by means of lake targets which comprise the volume (or level) and the frequency of occurrence of this volume/level.

The targets nominated by Southwell and Burns are listed in Table 1 of their report and reproduced in Table 2-2 of the CSIRO reports. There is some potential ambiguity about the target frequencies referred to by Southwell and Burns nevertheless the reviewer has assumed that the targets were intended to be the frequency of the maximum storage volume that occurs over a water year, which also appears consistent with the approach adopted by CSIRO13.

Achievement of these targets under the 'no impacts' scenario does not appear to have been reported by CSIRO. In order to investigate the performance of Lakes Menindee and Cawndilla, the reviewer obtained the time series of Lake volumes from CSIRO and has plotted the frequency in Figures 2 and 314. These figures indicate that both Lakes are proposed to be operated at similar or marginally lower levels than suggested by Southwell and Burns.

Alterations to the proposed operation that might assist in better achieving the targets could include a combination of:

(a) reducing outflow rates from the Menindee regulator;

(b) increasing the height of Lake Pamamaroo (and Wetherell) at the time when transfers are being made to Lake Menindee (and Cawndilla), so as to increase the volume of the transfers; and

(c) allowing the transfers to occur for longer.

If the lake operations were changed to better match the target environmental filling rules then evaporation will increase and water savings will reduce although it is unclear whether the resultant changes in water savings would be significant or not.

An analysis of the sensitivity of the water savings to achievement of the environmental targets would have been a useful addition to the CSIRO reports.

2.6 CSIRO DOCUMENTATION PROVIDED FOR REVIEW

CSIRO's documentation has been produced via the three reports prepared during November and December 2011. The latter two reports include refinements and supplements to the first and second reports.

Whilst this format of the documentation was no doubt dictated by the work program at the time, this format makes it difficult for a reader. In particular the final 'No Impacts' report is not a stand-alone document and the reader has to refer to the previous two reports in order to fully understand the basis of the modelling investigations being reported. Further it is not always clear which components of the first two reports have been superseded by the later investigations in the 'No Impacts' report.

13 The reviewer notes that the frequency of the maximum storage volume achieved in a year may in some cases be significantly different from the frequency of daily (or end of month) storage volumes. 14 Similar data was also obtained for Wetherell, Pamamaroo, Hume and Dartmouth. Frequency plots for these storages are provided in Appendix B. These plots show the frequency of achieving monthly volumes rather than the annual maximum frequencies depicted in Figure 2 and 3.


Figure 2: Lake Menindee Storage Frequency using data sourced from the CSIRO 'No Impacts' Report

Figure 3: Lake Cawndilla Storage Frequency using data sourced from the CSIRO 'No Impacts' Report


2.7 SECURING BROKEN HILL'S WATER SUPPLY

The Menindee Lakes System serves a vital role in securing a water supply to Broken Hill during periods of drought. However ever since construction of the Scheme in the 1960s it has been recognised that to provide this security from the Lakes, very high evaporation penalties are incurred. Consequently the MOU recognised that "a precursor to achieving these savings is the provision of an alternate, secure water supply for Broken Hill ..".

The reviewer understands based on his discussions with NOW and CSIRO that as at December 2010:

(a) Geoscience Australia had been engaged to undertake investigations into providing a secure water supply to Broken Hill utilising either:

(i) groundwater extractions (from a borefield near Menindee); or (ii) conjunctive use supply utilising a managed aquifer recharge (MAR) groundwater

scheme. Under this arrangement Broken Hill's supply would continue to be provided from the Lakes and only during drought periods would supplies be drawn from the aquifer. Nevertheless aquifer recharge would occur outside drought periods when the available water supply was of suitable quality;

(b) there was still a significant amount of work to be done before the conceptual design of the scheme could be developed and costed. The reviewer considers that a conjunctive use MAR scheme may possibly be favoured;

(c) the surface water demands required for the conjunctive use MAR scheme were unknown;

(d) CSIRO's modelling had assumed the groundwater extraction scheme (without recharge) would be implemented, i.e. the current practice of drawing water supplies from the Lakes would be halted. Consequently the 'base case' model diversions for Broken Hill (approx 10GL/yr) would contribute to the 'savings' identified in the 'no impacts' scenario. For the modelling to properly represent the conjunctive use scheme not only would these water supplies need to be re-instated but further extractions to simulate the aquifer recharge would also need to be included. (Nevertheless the magnitude of these changes is likely to be relatively small compared with the savings).

2.8 CAPACITY TO WITHSTAND DROUGHTS WORSE THAN HISTORICAL

At the onset of a drought when the Lakes revert to NSW control, there is potential for the water savings scheme to reduce the water security of water users at Menindee and along the Anabranch and the Lower Darling.

Under the existing arrangements, 480GL of water would be held in Lakes Wetherell, Pamamaroo, Menindee and Cawndilla. Under the proposed arrangements, 180GL will be held in these Lakes although the distribution of water between the Lakes will be different. The proportion of water will likely be higher in the upper Lakes, Pamamaroo and Wetherell, which may significantly alter the operational efficiency of the Lakes over the ensuing drought.


The capacity of the Lakes scheme to perform during droughts is likely to be a key issue for water users and the local community. In the opinion of the reviewer, the project must be able to demonstrate that the drought capacity has not been diminished and preferably has been extended, otherwise there will likely be significant community opposition to the scheme despite the water savings generated.

Both NOW and CSIRO have given consideration to this issue. NOW's Andrew Davidson and Andrew Brown produced a discussion paper on this issue on 14 December 2010 − refer Section 1.4.1(l) above. However the discussion paper raises matters that do not appear to have been addressed in the CSIRO 'No Impacts' report.15 The matter is complicated because historically the Lakes scheme has managed to pull through the worst droughts experienced since construction in the 1960s, and model simulations using data from the last 100-120 years indicate similar behaviour in the most severe historical droughts in the early part of the 1900s. Consequently the model analyses that have been undertaken to investigate this issue have used statistically generated streamflow data containing inflow sequences to the Lakes which are drier than those that have been observed. This has led to differences of opinion between the parties concerning which synthetic inflow sequences should be used and the probability of these flow sequences occurring.

CSIRO's analyses have shown that adverse impacts will occur at the start of the irrigation season but that there is opportunity for additional diversions to occur later in the seasons so that the total diversion over the season is not reduced. The reviewer agrees with NOW's comments that whilst this is technically possible, most irrigators planting decisions will be made early in the season and the shortage of water at that time will adversely impact their agricultural production. The subsequent availability of additional water later in the season will do little to offset this impact.

Further, the reviewer considers that the differences of opinion concerning the synthetic inflow sequences have diverted attention away from examining whether the drought capacity of the Lakes will be impacted adversely. This is a key issue which does not appear to have been properly investigated in CSIRO's documentation.

The reviewer questions whether it would be simpler to test the impacts on drought capacity through use of a zero inflow sequence. The key measure in such an analysis would be the duration of the supply until the 'crash'. In such an analysis, the water savings proposal would need to demonstrate that it lengthened the period of supply. Otherwise it could be inferred that the scheme was reducing the drought capacity of the Lakes.

The reviewer also notes that the project may cause impacts to South Australia under extreme low inflow conditions – refer Section 2.1.2(f)(iii).

2.9 FLOODING IMPACTS

The reviewer understands that flows in excess of about 22,000ML/d in the River cause flood problems at Menindee.

CSIRO have reported that:

(a) flows of this magnitude currently occur in 22% of years; (b) the frequency of this flooding will increase under the water savings scenario16;

15 It may have been that Davidson's and Brown's discussion paper was prepared subsequent to the CSIRO report. 16 The November 2010 'Options for Environmental Filling' version of the CSIRO documents stated that under the operation envisaged at that time, the frequency of flooding would increase to 28%. Whilst the water savings


(c) the frequency of flooding could be reduced to below current levels through changed operations of the Lakes but this would incur an additional evaporation penalty of about 11GL/yr; and

(d) a cost benefit study could be conducted to identify the merits of purchasing the flood affected properties in lieu of foregoing the evaporation penalty.

Based on the reviewer's discussions with NOW there does not appear to be any dispute about the magnitude of the adverse impact or the need for the project to address this issue in the manner suggested by CSIRO.

Compulsory acquisition of the flood affected properties will likely incur other social impacts which would need to be carefully considered before this mitigation option was pursued.

2.10 IMPACTS ACKNOWLEDGED IN CSIRO'S DOCUMENTATION

The following comprise negative impacts (of any size) that CSIRO acknowledges will occur:

(a) NSW Murray − mean November allocations are 1% less but these allocations recover by June, and there is a small deterioration in some general security allocations;

(b) flows to SA − there is a 1% increase in the number of years when these flows are below entitlement;

(c) the 95 percentile Morgan salinities increase by 17EC;

(d) there is a 9% increase in the number of months when the combined volume of Lakes Wetherell and Pamamaroo are less than 100GL;

(e) the drawdown of Lakes Wetherell and Pamamaroo during dry periods is greater;

(f) there is a small deterioration in some Lower Darling General Security allocations;

(g) there is a small deterioration in some Victorian Murray allocations for low reliability water shares (LRWS);

(h) there is a small reduction in the frequency mid-range storage volumes in Hume and Dartmouth17;

(i) there will be an exacerbation of existing flooding problems at Menindee unless mitigation measures are provided;

(j) under climate change with 'dry 2030' conditions generally the magnitude of the above impacts increases.

proposal reported in the 'No Impacts' report in December 2010 is slightly different from the November 2010 report, the reviewer was not able to identify any revised flooding impact assessment (and has consequently assumed it remains similar to that identified in the November 2010 report). 17 Although frequency distributions of the Hume and Dartmouth storage volumes are not provided in the CSIRO reports, time series plots are provided in Appendix B. Frequency analysis of this information indicates that mid-range storage volumes (around 50% of capacity) will occur slightly less frequently. A review of allocations (if available) would likely also show a small deterioration during periods when Hume and Dartmouth storages are approximately half full.


CSIRO have described all the impacts as "very minor" and either "within the error bounds of the model" or "can be offset". Some commentary on each of these terms is warranted:

(a) "very minor" − it appears that the magnitude of the impacts is not in dispute, rather it is the significance of the impacts which is in dispute between CSIRO and NOW. Based on discussions with officers from NOW, the reviewer's understanding is that should the project proceed, NSW will be responsible for obtaining all the necessary approvals including environmental approvals. Further as the reviewer understands that NOW has concerns that even small impacts may not be considered to be minor by the affected individuals and consequently they may be difficult to justify during the project approval process;

(b) "within the error bounds of the model" − the reviewer understands that the magnitudes of simulation errors within MSM-BigMod have not been quantified and would likely be very difficult to quantify. Such errors, often referred to as 'absolute errors' would measure the difference between the model predictions and reality. The reviewer agrees that these absolute errors would likely be larger or much larger than any of the impacts predicted for the 'no impacts' scenario. What is of interest here however is the 'relative errors' between the 'base case' and the 'no impacts' scenario runs utilised to determine the impacts. As is the model is deterministic, the reviewer anticipates the relative error would be very much smaller than the absolute error and despite the impacts being of small magnitude it is likely to be inappropriate to dismiss them on the basis that they are "within the error bounds of the model"; and

(c) "can be offset" − the reviewer agrees that a number of the impacts (e.g. increased Morgan salinities, reduced NSW Murray diversions) could be offset. It would seem appropriate for these offsetting mechanisms to be identified and costed. Indeed if the interpretation of 'no impacts' is as set out in Appendix A, the reviewer understands that NOW's position is that the MOU requires all impacts to be offset, or reduced to a level where it could be confident that the project would obtain all necessary environmental approvals without difficulty.

2.11 OTHER IMPACTS OR POTENTIAL IMPACTS OF CONCERN TO NOW

The following is a list NOW's concerns of impacts or potential impacts in addition to those listed in Section 2.10. This list has been prepared by the reviewer based on his discussions with NOW and his review of documentation provided by NOW:

(a) more frequent inundation of Lakes Wetherell and Pamamaroo. This would involve destruction of some of the vegetation surrounding Lake Pamamaroo and the upstream environment in Lake Wetherell;

(b) the high fills of lakes Menindee and Cawndilla are of insufficient frequency and magnitude to inundate Lake Speculation and sustain riparian values on the fringes of all lakes;

(c) the proposed size of the enlarged Menindee outlet regulator (14,400ML/d) may be too large. Consideration needs to be given to cultural values in the late bed, the area


disturbed during construction and potential impacts downstream from increased discharges;

(d) changes to the magnitude and frequency of flows along the Darling Anabranch will occur. The new flow regime along the Anabranch needs to be explicitly identified in the modelling and secured expressly for the Anabranch;

(e) new shoreline erosion issues may emerge;

(f) extended dry inflows sequences may exacerbate impacts of the proposal on consumptive users;

(g) impacts of changes at the Lakes in marginal years have not been rigorously considered in the CSIRO reports as the emphasis has been on average annual impacts;

(h) many of the potential social, economic and environmental impacts of the proposal have not been documented.

The reviewer understands that the NSW Government favours modifications to the proposal to include construction of a regulator at Morton Boolka which would allow Lakes Menindee and Cawndilla to be managed separately. It would also like to see the Penellco channel upgraded and operated as proposed by the Darling River Water Savings Project, with the environmental filling proposals restricted to Lake Cawndilla. They also wish greater emphasis be given to maintaining the social amenity of Lake Menindee.

2.12 PROPOSED USE OF SAVED WATER AND ITS PROPOSED LICENCE

The 'No Impacts' report discusses a potential licence at Weir 32 with conditions which mimic a NSW supplementary licence. These conditions comprise:

(a) maximum rate of extraction of 25,000ML/d;

(b) extraction only when Lake Victoria is spilling and spills at Weir 32 exceed 1,000ML/d; and

(c) annual diversion limit of 347GL.

The reviewer understands that this extractive licence was modelled and the resultant impacts were presented in Table 4-1 of the 'No Impacts' report.

This information compares directly with Table 2-3 of the report which reports the modelling of the 'operation rules' (i.e. without considering any extraction). In particular under these rules:

(a) water remains in the system and the majority passes downstream to SA (and some could be re-regulated in Lake Victoria if space is available);

(b) once an environmental filling target is met in Lakes Menindee and Cawndilla, water is drained at the maximum rate possible; and

(c) drainage of Menindee and Cawndilla occurs via the enlarged Menindee outlet regulator and through releases from Lake Cawndilla to the Anabranch.


Further the reviewer considers it likely that the limits specified in the licence conditions referred to above were determined by modelling to ensure the average annual flow to SA was not reduced below the base case level. A comparison of Table 2-3 and 4-1 shows the impacts are very similar although Morgan salinities have increased (likely due to the extraction of flows from the system). CSIRO make it clear that "the CEWH will not be taking water out of the system" and therefore the impacts to South Australia shown in Table 4-1 are over stated and will be more like those shown on Table 2-3. Whilst the CEWH might use some water for watering of a wetland (i.e. similar to an extractive use) rather passing it down to the Murray mouth, it would be difficult to see all the water used this way and therefore CSIRO's preference to use the Table 2-3 in lieu of Table 4-1 is probably justified18.

The magnitude of the supplementary type licence that is specified above is intrinsically related to the licence conditions particularly relating to the timing and location of the access. In a system like Menindee Lakes, changes to the timing or location may have significant implications on the resultant quantum of water that is available.

18 Nevertheless the reviewer understands the CEWH is not obligated to use its water in any fashion and future changing priorities could see the SA impacts change from either of these model outcomes. If some conditions were placed on the use of CEWH water then there would be more certainty about the downstream impacts.


3. REVIEW FINDINGS

3.1 TOR 1: ADEQUACY OF THE MODELLING ASSUMPTIONS

In particular report on the adequacy of the model assumptions and model settings used to: maximise evaporation savings; maximise entitlement that might be created from the evaporation savings; and minimise downstream impacts.

Response

(a) All assessments of the type undertaken typically utilise the MSM-BigMod modelling platform. This model has been peer reviewed and is an appropriate model for use in the circumstances under which it has been applied by CSIRO.

(b) It has not been possible for the reviewer to inspect all the modelling assumptions utilised in the 'base case' and the 'no impacts' simulations. The key assumptions are those which have been changed between the two simulations and, in the opinion of the reviewer, these assumptions are adequate for the purpose to which they have been used, subject to the qualification given below.

(c) The evaporation savings obtained will be strongly correlated with the environmental filling rules that have been imposed on the Lakes operations. The filling of Lakes Menindee and Cawndilla that has been achieved in the model is illustrated in Figures 2and 3. This filling appears to be of a marginally lesser extent than envisaged in the environmental filling rules developed by Southwell and Burns for the MDBA. It is unclear whether increases in the magnitude or duration of filling within the model to more closely match the environmental filling rules will lead to a significant reduction in the savings. Further model runs to test the sensitivity of the savings would be useful.

(d) CSIRO have performed model simulations which provide for the water used for the environmental filling of Lakes Menindee and Cawndilla, to be subsequently released to the Darling River (via the enlarged Menindee outlet regulator) or to the Anabranch (via the Cawndilla outlet regulator) at the maximum rate possible. Any reduction in the ability to drain these Lakes (e.g. through reduced outlet capacity from that assumed in the model) will increase the duration of ponded water in the Lakes and reduce savings. The reviewer confirms that the model assumptions used by CSIRO for the Menindee and Cawndilla regulator capacity is consistent with the Water Savings Project proposal.

3.2 TOR 2: REVIEW CONCLUSIONS FROM THE CSIRO REPORTS

In particular whether: all the assessment criteria have been addressed; the criteria of the MOU have been achieved; and the data would support any alternative conclusions.

Response

(a) There appears to be a disagreement about the assessment criteria that were adopted by the HMWG. NOW's assessment criteria, which the reviewer understands were


presented to the HMWG, are reproduced in Appendix A. The assessment criteria used by CSIRO are only a subset of these. A direct consequence of this is that all of NOW's assessment criteria have not been addressed in the CSIRO documentation.

(b) In general terms, the MOU requires there be no adverse impact to the environment or third parties. However the MOU makes no comment about the magnitude or significance of the adverse impacts. The reviewer interprets this to mean that no adverse impacts, however small, were envisaged by the MOU.

(c) CSIRO have stated that adverse impacts will occur but have described them as being either minor, within the error bounds of the model, or capable of being offset. If the reviewer's interpretation of the requirements of the MOU expressed in the previous paragraph is correct, then the requirements of the MOU have not been met.

(d) Further the reviewer notes that there are a range of impacts other than those reported by CSIRO which remain a potential concern. For example the changed water level behaviour in Wetherell and Pamamaroo will produce higher frequencies of inundation over some height ranges (see Appendix B).

(e) The reviewer has also been asked whether "the data would support any alternative conclusions". This is a broad question and the scope of the data being referred to is unclear. Nevertheless it is the reviewer's opinion that:

(i) there are a number of adverse impacts resulting from the proposal, many of which are small;

(ii) further work is required to quantify some impacts which have not been reported in the CSIRO documentation;

(iii) whether the impacts are significant is a matter for further social, environmental and economic assessment and the advice of other experts. The reviewer is not qualified to identify whether the size of impacts identified by CSIRO would present a major difficulty in subsequent environmental approvals under NSW legislation. Nevertheless in respect of at least the impacts on NSW Murray irrigators and the impacts of some elevated water levels in Wetherell and Pamamaroo, the reviewer has concerns that these will likely cause approval difficulties;

(iv) the potential to identify mechanisms or alterations to the proposal to offset adverse impacts and as a consequence produce beneficial impacts, does not appear to have been explored.

3.3. TOR 3: REPORT ON THE IMPACTS AND RISKS OF THE PROPOSED OPTIONS MODELLED ON:

NSW water users; Broken Hill; the Darling Anabranch; the Lake Wetherell floodplain; and Victoria and South Australia.

Response

(a) The impacts on NSW water users, the Darling Anabranch, Victoria and South Australia have been documented by CSIRO and the reviewer agrees with the quantification of these impacts;


(b) In regard to Broken Hill water users, their future security will be dependent on the proposed groundwater scheme which is a matter for others to assess. Based on the limited information which the reviewer has seen about the groundwater proposals, there appears to be an excellent opportunity for Broken Hill users to receive better quality water and possibly higher security water.

(c) In regard to the Lake Wetherell floodplain, the CSIRO reports do not quantify the impacts in sufficient detail. The plots presented in Appendix B suggest there will be some increased duration of inundation. The environmental significance of this increase is a matter for others to determine.

3.4 TOR 4: COMMENT ON THE EFFECTIVENESS OF ANY MITIGATION MEASURES PROPOSED IN THE CSIRO REPORTS

Response

(a) The reviewer is unaware of any proposed 'mitigation measures' except for the proposal to use salinity credits to offset impacts at Morgan and the alteration to the proposed Lake operations to mitigate any flooding impacts.

(b) Increases in salinity provide an economic impost on downstream users that must be considered in evaluating the appropriateness of the project. Provided this assessment is undertaken, and the Commonwealth agrees to relinquish some salinity credits, the reviewer sees no problem with this mitigation measure.

(c) It will likely be unacceptable to allow the project to exacerbate the existing flooding problems at Menindee. CSIRO have identified that it is possible to address this impact. Further investigation of alternative mitigation measures is required so the project can be modified to eliminate this impact.

3.5 TOR 5: REVIEW THE RISK ASSOCIATED WITH THE PREFERRED OPTION MODELLED

In particular, with respect to prolonged dry sequences and the ability to maintain supply to NSW Murray and Lower Darling irrigators.

Response

(a) Small adverse impacts to NSW and Lower Darling irrigators will occur. The question of the significance of these impacts and the ability of the NSW Government to obtain environmental approvals for a water savings option with such adverse impacts is ultimately a matter for others to determine.

(b) Measures to offset these impacts and/or to provide a benefit rather than a dis-benefit to these water users do not appear to have been investigated.


3.6 TOR 6: COMMENT ON ISSUES RELATING TO THE FORM AND QUANTUM OF THE NEW ENTITLEMENT PROPOSED TO BE CREATED

In particular, how the future use of that entitlement may impact on different user groups depending on where the environmental water licence is utilised.

Response

(a) The reviewer understands a Supplementary type licence has been proposed at Weir 32 for the CEWH's water. Modelling of this licence has been undertaken assuming extraction at Weir 32 (i.e. Table 4-1 of CSIRO's 'No Impacts' report).

(b) Modelling has also been undertaken of the 'operation rules' without any extraction (i.e. Table 2-3 of CSIRO's 'No Impacts' report). Under this scenario the CEWH's water passes downstream and is not consumed.

(c) The two modelling scenarios described in (a) and (b) above provide for the water to be used either at Weir 32 or at the Murray mouth. These two scenarios provide an indication of the potential changes which could occur for different locations of the proposed use.

(d) For a location below Weir 32, a lesser entitlement volume would apply due to losses enroute. Nevertheless additional in-stream benefits19 will accrue between Weir 32 and the location of the use, due to the additional river flows.

(e) Further analysis of these issues would be assisted if the CEWH provided details of how it proposes to use the water.

(f) The quantum of the entitlement and the quantum of the water savings reported by CSIRO may also need to be adjusted for one or more of the following matters identified in this review:

(i) any changed operation to ameliorate the identified flooding impacts (refer Section 2.9);

(ii) any changed operation to better achieve the environmental filling targets (refer Section 2.5);

(iii) allowance for additional supplies to secure Broken Hill's water supply (refer Section 2.7); or

(iv) other measures that may subsequently be introduced to offset any other adverse impacts.

3.7 TOR 7: EXAMINE THE INTERACTIONS OF THE PROPOSED OPTION WITH CURRENT AND FUTURE BROKEN HILL WATER SUPPLY ARRANGEMENTS

In particular: how an MAR scheme might impact on Lake operations and hydrology; how BH could source additional entitlement to allow for aquifer losses and how this

might impact on Cap issues; and whether a surface storage solution with backup from groundwater might impact on Lake

operations and how this might impact on potential savings.

19 Some disbenefits, e.g. flooding, may also occur.


Response

(a) Based on discussions with CSIRO, the reviewer understands the modelling has not provided for any simulation of the existing Broken Hill water supplies as part of the 'no impacts' scenario. (The modelled situation would only be appropriate if the supply was to be taken entirely from groundwater without use of any surface waters).

(b) The proposed conjunctive use MAR scheme however will require a continuation of the existing supplies in normal years as well as the transfer of additional surface supplies into aquifer storage. Consequently the currently reported savings will be reduced once these two adjustments are made to the 'no impacts' model. The first adjustment will be about 10GL/yr (i.e. the current Broken Hill supply simulated in the 'base case' model). The reviewer understands the quantum of the second adjustment which will provide essentially for aquifer losses, is currently not known.

(c) The additional entitlement required to supply these aquifer losses and address cap issues would supposedly be addressed through a reduction in the savings generated by the project.

(d) The reviewer has also been asked to comment on "whether a surface storage solution with backup from groundwater might impact on Lake operations and how this might impact on potential savings". It would be inappropriate for the reviewer to make recommendations concerning the preferred method of supplying Broken Hill as the issues are complex and extend beyond hydrology (e.g. water quality, capital and operating costs). Nevertheless the reviewer can advise in terms of the hydrological impacts of the two schemes on the Lakes, the relative impact is directly related to:

(i) in the case of the MAR scheme, the magnitude of the aquifer loss which must be supplied by surface supplies; compared with

(ii) in the case of the surface storage scheme, the magnitude of the evaporation losses from the surface storage that is envisaged in TOR 7.

3.8 TOR 8: DISCUSS ALTERNATE OPTIONS

In particular whether Options B4 � B6 would provide better outcomes in terms of user and environmental impacts while still achieving significant water savings.

Response

(a) The reviewer has a limited knowledge of Options B4-B6, the resultant water savings and impacts.

(b) Conceptually these options involve less dramatic changes to the existing operation of the Lakes. Consequently they would likely provide generally less water savings and fewer impacts.

(c) The reviewer is unable to provide any further detailed response on the merits of these schemes without additional work which is beyond the scope of this review.


APPENDIX A

NOW's CRITERIA FOR 'NO IMPACTS' (This is Version 5 of the document prepared by Andrew Brown and Lindsay White

"Measuring No Downstream Impacts and Lake Failure" dated 15/9/2010).

Criterion highlighted in green are consistent with those reported in Table 1-1 or plotted in Appendix A of the CSIRO reports.

Criterion highlighted in yellow have not been reported numerically or discussed in any detail in the CSIRO reports. (Note that in a few cases CSIRO may have used

other measures that are not included in NOW's list of criteria).

Review of Menindee Lakes MOU Investigations Bewsher Consulting Pty Ltd Final Report, January 2012 31 J1952R_4

Measuring No Downstream Impacts and Lake Failure

Andrew Brown & Lindsay White– Version 5 – 15/9/2010

General Description

This is a 3 stage process that can be characterised as iteration, checking, and reporting

In the first stage (iteration), the modellers should use summary statistics provided on the standard MSM output spreadsheet to monitor the system behaviour during the modelling exercise. The key metrics are expected to be those listed in Table 1.

In the second stage (checking), the modellers would believe that they have achieved the purpose of the modelling exercise, and are carrying out a closer inspection of time series to look for unexpected behaviour. The object is not to identify individual examples of a minor impact occurring, but evidence of consistent trends. It also allows for identification of catastrophic events that can be masked by average style statistics. A suggested method is in Table 2

In the 3rd stage (reporting), more extensive outputs for a subset of model runs should be provided to a wider audience. The wider range of outputs is intended to address matters that are normally monitored by jurisdictions, but are not expected to be significantly impacted by this project.

Iteration Targets

Measure Purpose CommentLower DarlingLower Darling Average % GSEC allocation in November not lower

Timeliness indicator on GSEC allocations to see effect on annual crop planting decision

Check if allocation improvements are being delayed

Lower Darling Average % GSEC allocation in June not lower

Total GSEC allocations not reduced Direct impact

Lower Darling Average % HSEC allocation in June not lower

Total HSEC allocations not reduced Previously had default protection from reserves kept for Broken Hill. May need a specific reserve in future.

Longest period with Lower Darling % HSEC allocation less than 50% is not greater

Worst dry spell not extended.

Lower Darling Average Annual Tandou Diversions not lower

Diversions not reduced

Lower Darling Average Annual non-Tandou Diversions not lower

Diversions not reduced Separate out effects of Tandou inter-valley transfers

Lower Darling Local water Utility remains at 100% Not failing Broken Hill, Pooncarie and Menindee Will need to include consideration of MAR effects Anabranch Environment average annual volume not lower

Protect some Ana branch’s share of surplus flows Could go up if some CEWH water is directed down there. Could go down if Cawndilla is kept dry.


NSW MurrayNSW Murray Average % GSEC allocation in November not lower

Timeliness indicator on GSEC allocations Check if allocation improvements are being delayed

NSW Murray Average % GSEC allocation in June not lower

Total GSEC allocations not reduced Direct impact

NSW Murray Average Annual diversions not lower Diversions not reduced Direct impact NSW Murray HSEC allocations not less than 97% in any June.

Test if system has failed. Must provide full HSEC by end of water year.

VIC MurrayMinimum VIC Feb allocation not lower Minimum water availability measure for VIC at end

of main irrigation period Standard VIC measure

% of Years VIC allocation <100% not lower Measures degree of success of hold for next year policy

Standard VIC measure

Average Victorian allocation Measures general water availability Standard VIC measure Average Annual VIC diversions Measures effects on irrigators Standard VIC measure SA% of years SA restricted not higher Measures general effects on irrigators Maximum SA restriction not higher Measures the depth of dry periods Flows to SA not reduced Measures gross water that goes to SA Menindee & Cawndilla EnvironmentSome measure of success in meeting the watering target

Did we hit the target or not Cannot be determined until we know the watering regime required.

Lower Darling Environment/Riparian Uses% of months that Weir32 target flows are not reached is not greater

Measures the amount of time that a riparian style flow is being made down the Darling

The WSP contains a target flow at Weir32. No reserves are kept to supply this flow or high security users and the flow is discontinued if required to preserve water for Broken Hill. However, it is important for S&D supplies and general river health. The 640/480 rule provided a high degree of success in meeting this flow target. If the 640/480 rule is to be reduced, this flow target may need to be catered for directly. This might be done with a reserve or adequate NSW/MDBA sharing rule. For example, to provide 2 years security for high security & Lower Darling basic, needs the reserved water would need to be about 300 to 350 GL.

Longest duration that Weir32 target flows are reached is not greater

Measures the longest period that a riparian style flow is not being achieved


Table 1 – Summary Statistics

Checking Time Series

Time Series Visual Inspection GuideRanked plot of Lower Darling GSEC allocations in February & June– Looking at the relative slopes of the transition from 100% to 0%. We don’t want to see a significant steepening that indicates we are making the system more boom and bust in nature. Continuous or Histogram plot of Lower Darling GSEC allocations in February & June – Skim through the time series to identify key dry periods. We don’t want to see evidence of consistent significant lengthening caused by earlier failures and/or slower recoveries. Also consider comparative time spent at very low allocation levels. Repeat previous two graphs for NSW Murray, VIC Murray, SA. Continuous or Histogram plot of combined Menindee Scheme lakes and individual lakes. Looking for evidence that the lakes are not meeting target distributions of volumes between lakes, and more frequent or extended periods of very low levels. Continuous or Histogram plot of Hume & Dartmouth lakes. Looking for signs that the Murray system is being stressed over the long term. This would show as Dartmouth being slow to recover from drawdowns and Hume not following its typical annual cycle.

Table 2 – Time Series Checking

Standard Hydrological Modelling Indicators For Reporting

Hydrological indicators are statistical parameters calculated from the time series of a given variable the model outputs (e.g. mean annual flow to SA). The indicators provide an effective way to assess the impacts/benefits of a given modelled scenario when compared to the reference run conditions. The following sets of hydrological indicators will be used to assess impacts/benefits of the options. The total set of indicators is broken down into four tables. The indicators specific for each of the three jurisdictions has been included as a separate table. The fourth table lists indicators that would provide general information on the River Murray System.

Table lists the indicators which are used to provide information on NSW’s High and General Security allocations in June and November respectively as well as mean and minimum annual diversions. NSW indicators are provided for both the Murray Valley and the Lower Darling.


Table 3 - NSW Indicators Description Units NSW Allocations

Percentage of years NSW Murray High Security allocations < 100% (Jun) % Mean NSW Murray High Security allocation (Jun) % Mean NSW Murray General Security allocation (Nov) % Mean NSW Lower Darling General Security allocation (Nov) %

NSW DiversionsMean annual NSW Murray diversion (all months) GL/yr Minimum annual NSW Murray diversion (all months) GL/yr Mean annual NSW Lower Darling diversion (all months) GL/yr Minimum annual NSW Lower Darling diversion (all months) GL/yr

NSW Lower Darling System Months that the volume in the Menindee Scheme is < 100 GL % % time Scheme is NSW control , within 50 GL of threshold and NSW Murray GS allocation less than 20% %

Months that Menindee and Cawndilla Lakes are full % Months that the Menindee scheme in NSW control %


The indicators chosen to show impacts to Victorian allocations and diversions are provided in Table 4. Indicators relating to Victoria’s high and Low Reliability Water Share allocations are provided for the end of February.

Table 4 - Victorian Indicators Description Units Victorian Allocations

Percentage of years Vic High Reliability Water Share < 100% (Feb) % Mean Vic Feb High Reliability Water Share (Feb) GL/yr Min Vic Feb High Reliability Water Share (Feb) GL/yr Percentage of years Vic Low Reliability Water Share < 100% (Feb) % Mean Vic Feb Low Reliability Water Share (Feb) GL/yr

Victorian DiversionsMean annual Vic Murray diversion (all months) GL/yr Minimum annual Vic Murray diversion (all months) GL/yr

Table 5 lists the South Australian indicators that are used to assess impacts/benefits. The restrictions to entitlement are given as total restriction in GL for the year. The minimum annual flow to SA is important indicator of the security of dilution and loss supply to SA and the provision of CHWN.

Table 5 - South Australian Indicators Description Units SA�s Entitlement Flows

Percentage of years SA entitlement restricted (all months) % Maximum annual SA restriction (all months) GL/yr Mean annual SA restriction (all months) GL/yr Minimum annual SA flow (all months) GL/yr

SA Diversions Mean annual SA Murray diversion (all months) GL/yr Minimum annual SA Murray diversion (all months) GL/yr


A number of general indicators are used to assess impacts to the overall River Murray System. A list of these indicators is provided in Table and is broken up into a number of categories.

Table 6 - General River Murray System Indicators Description Units General Flow Indicators

Mean annual flow downstream of Doctors Point (all months) GL/Yr Mean annual flow downstream of Yarrawonga (all months) GL/yr Mean annual flow downstream of Euston (all months) GL/yr Mean annual Burtundy flow (all months) GL/yr Mean Darling Anabranch outflow (all months) GL/yr Mean annual flow to SA (all months) GL/yr Mean annual flow at Lock 1 (all months) GL/yr Mean annual flow over Barrages (all months) GL/yr

Loss and Spill Indicators Mean annual Total Murray System loss (all months) GL/yr Mean annual Total Darling loss (all months) GL/yr Mean annual Total Murray evaporation loss (all months) GL/yr Mean annual Murray System spill (all months) GL/yr

Level Indicators Mean Lower Lakes Level (all months) mAHD Minimum Lower Lakes Level (all months) mAHD

General Salinity Indicators Mean daily Morgan Salinity (all months) EC 95 percentile daily Morgan Salinity (all months) EC Mean daily Burtundy Salinity (all months) EC 95 percentile daily Burtundy Salinity (all months) EC Mean daily Murray Bridge Salinity (all months) EC % of days the Salinity at Murray Bridge is > 1400 EC %


A broad range of general flow indicators were chosen to give an indication of possible changes in flow regime throughout the River Murray System and provide an assessment of the total volume of water moving through the system. In addition, the last indicator (mean annual flow over the Barrages) provides a measure of how much water is reaching the end of the system.

Four loss and spill indicators were chosen being the mean annual loss from the Darling and the Murray Rivers; the evaporation loss from the Murray System (calculated as the sum of the evaporation loss from Dartmouth Reservoir, Hume Reservoir, Lake Victoria and the Menindee Lakes Storage); and the mean annual volume spilt from the main Murray System storages. These indicators provide an indication of the volume of water being stored and lost from the regulated storages on the Murray System. The overall spill indicator has been included to provide information about possible changes to spills including unregulated flows.

In addition to the regulated storage indicators two level indicators are specifically for the Lower Lakes to gauge the effect a possible change will have on the terminal lake system.

The salinity indicators are the average daily salinity at Morgan and the 95th percentile daily salinity at Morgan which relate to targets set in the BSMS program. The salinity at Burtundy provides information about salinity of water coming into the system from the Darling, while the salinity at Murray Bridge is a good indicator for the Lower Murray impacts.


APPENDIX B

STORAGE FREQUENCY PLOTS (The source data for these plots was provided by CSIRO from the

modelling presented in their 'No Impacts' report. The base data represent the simulated end of month storage volumes).

