Transcript for Lake Eildon Technical Assessment Webinar

Beth Ashworth (DEECA)
Welcome everyone and thank you for making the time to join us today for the Lake Eildon Technical Assessment webinar. I'm Beth Ashworth. I'm Acting Executive Director of Water Resource Strategy at DEECA (Department of Energy, Environment and Climate Action), and I'll be introducing this session today. I'd like to start by acknowledging the Traditional Owners of the various lands that we're calling in from, in particular the Taungurung and Yorta Yorta nations, as the Traditional Owners of the lands and waters that we're here to talk about today. I pay my respects to their elders, past and present, and extend that respect to any elders we've got online with us today.

I'd also like to introduce Hench Wang and Simon Lang from HARC [Hydrology and Risk Consulting Pty Ltd], who undertook the assessment and are here to provide an overview and answer your questions today. But before proceeding with the webinar, I'd like to recognise and thank all the community members, volunteers, local councils, and agency staff who worked tirelessly to keep people safe during floods. I also recognise that one and a half years on from the October 2022 floods, there are many people who are still deeply affected and dealing with the ongoing challenges associated with that, and the subsequent floods that have occurred since then.

In February last year, the Minister for Water announced the technical assessment of operating arrangements at Lake Eildon, and today we're reporting back on what that assessment found. The assessment was initiated in response to community feedback specifically about Lake Eildon and it focuses on specific management options there. It doesn't have the broader scope that a full flood study or flood management plan will, that will be the next step.

First, I'll provide an overview of the scope, what was and wasn't included in the technical assessment, and give a brief overview of the findings and next steps. Then I'll hand over to Hench who will talk a little bit more in detail about the assessment itself and a more detailed explanation of the findings.

There is a Q and A button you'll find at the top of your screen to ask any questions, and we'll have plenty of time at the end to answer those. The questions will be moderated, which means our moderator won't be publishing any questions with foul language or the like. I completely appreciate the trauma and challenges that people who have been affected by the floods feel, but I ask that you remain respectful in your questions and keep them focused on the technical assessment and the information we're providing today. The webinar is being recorded and will be available on DEECA's website, along with a copy of the transcript.

Moving on to providing a little bit of information about the context for this project. The primary purpose of Victoria's water storages and of Lake Eildon is to provide secure and safe water supply whilst offering some flood mitigation where possible. Those water supplies ensure communities can get clean drinking water from their taps, they underpin much of the irrigated agriculture in the GMID [Goulburn Murray Irrigation District] and the economic prosperity of the region, as well as support the health of the region’s waterways as well. Now, despite that being the primary purpose of Lake Eildon, it does still provide flood mitigation benefits even under the current operating rules. For example, during the October 2022 flood event, peak inflows to Lake Eildon were about 145,000 megalitres per day and peak outflows from [Lake Eildon] were 38,000 megalitres per day. While we recognise that is still a high outflow, Lake Eildon did reduce the peak by more than 70 percent.

Following those 2022 floods, and in response to community feedback, DEECA engaged HARC to undertake an assessment to see if the operating arrangements at Lake Eildon could be changed - including whether that would help mitigate flooding downstream from Lake Eildon to Seymour and the financial and non-financial impacts of that. A working group involving DEECA, local councils, water corporations, Goulburn Broken CMA [Catchment Management Authority] and the Victorian Environmental Water Holder were involved in scoping and supporting the assessment and that resulted in six options being investigated.

The Victorian Floodplain Management Strategy sets out policy for investment in flood mitigation infrastructure. It includes an investment principle of cost effectiveness, where the flood mitigation benefits must be greater than the total cost, including the capital and ongoing costs. So, a high-level cost benefit analysis was a really important part of this assessment. The final report is now public and available on DEECA's website.

From the outset, I just want to be really clear on what was included and not included in the assessment. It considers the initial cost to offset any water reliability impacts to entitlement holders. That is because some of the options look at reducing the volume of water held in Lake Eildon and that would affect the reliability of entitlements. HARC looked at how much water would need to be recovered to offset that [volume] and estimated the cost associated with that. We refer to that as the initial capital cost.

Secondly, the assessment calculates the tangible flood damages from Lake Eildon to Seymour, and that includes building and contents, vehicles, roads and rail, agriculture impacts of the peak flow, as well as indirect costs. For each of the options, the assessment looks at how the costs would change or reduce if the option was implemented, and those avoided flood damages are what we refer to in the assessment or what's referred to in the assessment as the benefits. We know that there are some costs missing and we know that there are some benefits missing. We know this isn't an exhaustive cost-benefit analysis and that there are factors on both sides that aren't accounted for.

[On the cost side,] the assessment doesn't consider the ongoing socio and economic impacts of reducing the volume of entitlement held in the Goulburn System. It [also] doesn't include the costs of modifying community assets around Lake Eildon, for example, the boat ramps, recreational facilities, etc. Or the cost of offsetting the reduced income to Goulburn-Murray Water from the fees that are associated with storing water, if there are entitlements retired from the system. And that’s really important that those costs are offset and so that fees for the remaining customers aren't increased.

Then on the [other] side, we know the assessment doesn't consider the [reduction in] intangible flood damages such as the mental health costs, the social welfare costs (we know those are important and often a dollar value can't be put on those), other tangible damages such as private recovery costs, environmental recovery costs and agricultural impacts from the duration of the flooding. But it's worth noting on that benefit side, there has been a 30 percent uplift to the [reduction in] tangible damages that's been included and that would go some way to accounting for some of those benefits.

The assessment also hasn't looked at the impacts to Traditional Owners, which could be either negative or positive impacts.

So, moving on to summarising the findings of the assessment, all options that were explored were found to be either ineffective or unreliable as flood mitigation strategies or they created costs and impacts that outweighed the flood mitigation benefits. Therefore, the assessment has found that none of the six options warrant further assessment. It also found that the tributaries that flow into the Goulburn system downstream of Lake Eildon, rather than the lake’s releases, have a much more significant impact on flooding in some of those downstream towns, like Molesworth and Seymour.

We know this is not the outcome of the assessment that many of you would have been hoping for and after my conversations with many flood-affected community members over the past year, I would personally have liked the assessment to find an option to pursue further. But unfortunately, the evidence isn't there to support that. I'm sure many of you have got questions about that, so after HARC provides some further details on these findings, we'll have plenty of time to answer those.

But first, I just wanted to touch on the next steps, because the findings of this assessment are far from the end of the road in terms of preparedness for future floods. The Victorian Floodplain Management Strategy sets out the roles and responsibilities in terms of floodplain management. Local councils lead, with support from their local catchment management authorities, flood studies in their local areas. That is to ensure that we've got locally-led solutions that meet local needs and include local communities. This technical assessment is available to local councils as an input into their flood studies and to Goulburn Broken CMA as well. Those flood studies aim to enhance community awareness and preparedness regarding flood risks at the local level.

We know flood studies are really important and the Victorian Government remains committed to supporting action by local councils in your area. In addition to the well-established programs, such as the Risk and Resilience and Disaster Ready Fund, the Victorian Government's committed an extra $10 million over a five-year period for flood studies in regional Victoria. Local councils can access that funding by working directly with their local CMA.

That was just a little touch on the next steps, but now I'll hand over to Hench from HARC, who will take you through in more detail the findings of the assessment. Thanks, Hench.

Hench Wang (HARC)
Thanks for that Beth. I just wanted to welcome everyone who's dialling in today to view this webinar. I wanted to provide an overview of the technical assessment that we've been engaged by DEECA to undertake over the last six to seven months. I will present on some of the findings as a result of that technical assessment.

I'll first go over an overview of the scope of work that we've been engaged by DEECA to do over the last six to seven months and [this] includes looking at the flood frequency changes. So that's looking at the impact on the peak flows if the operating options were to be implemented. The key locations that we focused our assessment on is the Lake Eildon outflows, the flows at Molesworth and the flows at Seymour.

Further to this, we've also looked at if the 1993 and 2022 floods were to be repeated, and what impact that may have on the peak flows estimated at those locations as well. We've also looked at the water resource implications associated with the implementation of the options as well as the recreational impacts around Lake Eildon. For example, if the water bodies level was to be lowered, what recreational impacts could that have. We also looked at the daily flow changes, and using the flood frequency changes in the first dot point and the water resource implications in the third dot point, look to tie those together to look at the avoided flood damages versus initial capital cost ratios. And from that some findings have been made around the benefit-cost ratios for the different options.

In terms of the area of assessment, we've focused our assessment on looking at the impacts to the flood frequencies, as well as the flood damages downstream of Lake Eildon to Seymour. That's shown on the region in the orange as you can see here. Within that region, the three key locations that we've looked at is Lake Eildon, downstream of Lake Eildon, Molesworth and Seymour. Later on, in some of the slides presenting some flood damage estimates, these are referred to as Eildon, Molesworth and Seymour. However, they cover the different reaches of the Goulburn River floodplain, that includes Lake Eildon to upstream of Molesworth, referred to as Eildon. Molesworth refers to Molesworth to Seymour, and then Seymour is the township of Seymour.

I'll also be covering off on some of the key terminology that will be referring to as part of today's presentation. The first is FSL and what that refers to is the full supply level or 100 percent capacity at Lake Eildon. The other concept is the flood mitigation provided. Currently Lake Eildon provides a level of flood mitigation, however, the options that we've investigated as part of this technical assessment look to provide additional flood mitigation. So, what is the additional incremental flood mitigation you could have with these options in place?

The third is surcharge, so that's holding water in Lake Eildon above 100 percent capacity or full supply level or FSL and releasing it after the flood has passed. And what that looks to do is to reduce the peak from the dam, by holding more water in the dam during a flood and letting that go after the flood peak has passed.

The fourth is the TPI, which is the Tripole index, which combines the influence of both the El Nino-Southern Oscillation Index, and I'll refer to it as ENSO during this presentation, and the Indian Ocean Dipole or IOD on rainfall across Australia. It's one of the key climate signals that we've been using to explore the effectiveness or the reliability of implementing climate signals for reducing the target storage.

The other concept that I'll be talking about is the target filling curve. What that is a scenario that's being adopted as the base case scenario for the modelling and an example of the target filling curve that was adopted for the 2023 calendar year as shown on the right. On the Y axis, you can see the volume of the Lake Eildon storage and on the [X] axis is the monthly date. So, by 1st of May the aim of the target filling curve is to fill Lake Eildon to 93.1 percent FSL and as the months move on, by the 1st of November it's reaching 100 percent capacity or full supply level. These target filling points are set so that Lake Eildon will reach 100 percent capacity within 95 percent probability of exceedance of inflows. It's also important to recognise that the target filling curve is adaptable and flexible, and they get updated throughout the season based on Bureau [of Meteorology]’s latest forecast and expected demands.

Another concept will be referring to in this presentation is the downstream flow constraints and that is the operational releases from Lake Eildon depend on the downstream flow constraints. These are 9,500 megalitres a day at Lake Eildon and 10,000 megalitres a day at Molesworth in the mid-Goulburn. It's important to also note that there's currently no stream flow or operational stream flow gauge at Molesworth. The purpose of these downstream flow constraints is to limit the flow within the Goulburn River channel. Later on in in some of the slides, I'll be referring to these in a fair bit of detail to better describe the impacts that downstream flow constraints have on the ability to provide air space all year round.

As a broad overview for the options that we've assessed, they've been developed through a Working Group that was at a workshop in September 2023. The six options are presented here on this slide. The first is changing the target filling curve., that looks at changing the percentage of exceedance to define the inflows for the target filling curves or looking at the delaying the target filling. Currently it can be 1st of October or 1st of November, but this is looking to delay that target filling to be 1st of December or 1st of January. The other aspect that's been looked at with the target filling curves, is also changing the period or the climate used to define the filling curves.

The second option is looking to reduce the target storage at Lake Eildon to 78 [percent], 85 [percent], 90 [percent] or 95 percent of full supply level. That option looks to provide more airspace in Lake Eildon all year round to have some flood mitigation benefits.

The third option looks at reducing the target storage based on using climate signals that indicate wet conditions. As part of this, we've explored a range of different climate signals which include using the ENSO index which defines periods of El Nino or La Nina which are drier or wetter conditions. Also looked at the Indian Ocean Dipole which can be negative or positive as well as using the TPI index which is a combination of both ENSO and the IOD indices.

For the fourth option, it looks to increase [or] have higher or more significant pre-releases from Lake Eildon based on forecast rainfall. So that's being able to provide additional airspace within the dam in the lead up to a flood event that is forecast to occur within the Lake Eildon catchment.

The fifth option looks to change the maximum surcharge. Currently Lake Eildon can be surcharged up to 600 millimetres. However, that option is looking to increase the surcharge capacity at Lake Eildon to 900 millimetres or up to 1.2 metres at Lake Eildon. What that [surcharge increase] aims to do is to hold more volume back during a flood event and then it is released after the flood peak has passed.

Option number six is looking to restrict the maximum outflow. If the outflow at Lake Eildon was restricted to minor, moderate or major flood class levels and the dam was allowed to surcharge more than its current capacity, what impact could that have in terms of flood mitigation benefits?

Something that I've highlighted on this slide is that of the six options that looked at this, the options three to six were found to not be robust ways to improve or increase the flood mitigation at Lake Eildon. I'll go into more detail in the subsequent slides about why options three to six aren't robust at providing additional flooding mitigation.

To put this in a little bit of context, particularly around option three which is why this target storage can't be reduced by using the climate signals. The recorded streamflow data downstream of Lake Eildon, Trawool and Seymour has been plotted on this graph. This has been plotted post-1975 with the flows on the Y axis, which are megalitres per day. Based on the recorded stream flow data, September 1975, October 1993 and October 2022 were the three largest flood events within that period of time within the Goulburn River system between Lake Eildon and Seymour.

In order to provide a better understanding of how effective is option three, using climate signals to reduce the target storage for some of those events, we've looked into the water resource modelling. This is to get a better understanding of how effective that option three is. To provide a context for this option, this option is using the TPI or the Tripole index as the climate indicator, and a value of less than negative one point five (-1.5) [trigger one] has been adopted to indicate a wet year. If the climate signals had a value less than that, it means that within the water resources modelling, it aimed to reduce the target storage to 85 percent of the full supply level [trigger two]. In order to for this option to be effective, it needs both trigger one and trigger two to be met in order to provide some flood mitigation benefits by the way of airspace within Lake Eildon.

We've implemented this within the water resource modelling and found that across the model’s full records there are only 20 months where these two triggers are met. Out of the 120 years of data, there's only a very small handful of months in which the conditions of trigger one and trigger two are met. There are only a very small number of months within the water resource modelling where the model attempts to reduce the Lake Eildon storage to 85 percent full. If they are met, based on some of the case studies that I highlighted in the previous slide, the downstream flow constraints do limit the ability to reduce the target storage down to 85 percent.  Also, they aren't reliable measures of being able to predict wet conditions and I'll provide more context within the next slides.

But for instance, the September 1975 and the October 2022 floods, the downstream flow constraints limited the ability to provide enough airspace. Even though in those years the climate signals were indicating that there were wet conditions, and the storage levels were above the target storage. However, the downstream flow constraints limit the ability for additional airspace to be provided. The other two flood events of May 1974 occurred during conditions where rainfall fell downstream of the dam and Lake Eildon wasn't above 85 percent for that flood event. Then for the 1993 flood event, which was the flood of record out of Lake Eildon, that was during El Nino or dry conditions years. So, if even if option three was implemented during 1993, there would have been no impact. I'll provide in the next few slides some more details summarising some of the key aspects around what the modelling showed for option three.

On this slide here, on the first graph [top] there is just the differences in the storage levels within Lake Eildon during the mid-1970s. We can see from this example in the mid-1970s that the option [three] was effective in being able to provide airspace, because the climate signals indicated wet conditions, which met that first trigger criteria, and also the target storage was above 85 percent. However, because most of the rainfall fell downstream during the 1974 event, there would not have been any differences with the 1974 event because all of the rain fell downstream. However, the 1975 event example, the downstream flow constraints limited the ability to be able to provide that additional airspace within Lake Eildon. From the outflow’s perspective, you can see that compared to the downstream flow constraints, that they were exceeded and therefore there were some spills that occurred. The impact on the outflows were still quite minor for the 1975 event. For those events in the mid-1970s, the climate signals did indicate that it was wet conditions. However, the effectiveness of the option [three] in being able to reduce the big outflows from Lake Eildon were limited, because of in the 1974 event, for example, the rainfall fell downstream of Lake Eildon and in the 1975 event it was limited because of the downstream flow constraints.

For the 2022 flood event, it was similar thing where the climate signals indicated wetter conditions, and so that meant the first trigger, and the target storage was also above 85 percent [trigger two]. However, similarly with the September 1975 flood event, the 2022 flood event only showed minor additional flood mitigation benefits as a result. This is because in the months leading up to the 2022 flood event, the downstream flow constraints limited the ability to provide enough airspace to significantly reduce the peak flows out of Lake Eildon.  

For 1993, as I mentioned earlier, the climate signals during 1993 and 1994 were during dry conditions. You can see on the figure on the bottom that it's plotting above the TPI value of 0. However, if that option [three] was implemented it would not have changed the impact of, or the outcomes, of the 1993 flood event.

In order to provide a summary for option three, the degree to which option three could be implemented during wet conditions was limited due to the downstream flow constraints. As highlighted with the 1993 event, the floods have occurred when the climate signals indicate drier or El Nino conditions. But also, the climate signals tested are generally poor predictors of the monthly inflows as well as the Lake Eildon storage levels. Therefore, option three was not a reliable or robust way to increase the flood mitigation provided by Lake Eildon.

As for option four, this is looking to provide more significant or increased pre-releases based on forecast rainfall. What this option looks to do is to create more airspace within the lake before the forecast rainfall events occurs, to accommodate the predicted inflows. With rainfall forecast products that are currently available, they're not reliable for an extended period where [pre] releases are required to provide that significant airspace. Rainfall forecasts need to be reliable more than two days in advance to be able to provide some of that airspace for flood mitigation. There's also a significant amount of uncertainty associated with forecast rainfall in where that will occur and whether or not the rain falls upstream of the dam or downstream of the dam, like in the 1974 and October 2022 flood events. There's a significant amount of uncertainty associated with them. Also, the floods have the potential, depending on whether the rainfall falls upstream or downstream of the dam, to be exacerbated by more significant pre-releases from Lake Eildon. As a result, option four was not found to be a reliable or robust way to increase the flood mitigation provided by Lake Eildon.

For options five and six, these were similar options in that they increased the surcharge, or the volume of water held within Lake Eildon during a flood event. For option five, that looks to increase the maximum capacity at which surcharge can occur, so [the lake] holds a greater volume. Option six looks to restrict the outflows. Whether or not that outflow is contained to match the downstream flood class levels of minor, moderate or major flooding or is being restricted to a lower outflow, it does surcharge the storage more [than current operations]. Therefore, after the flood event a greater volume of water will need to be released after the flood peak has passed, which potentially elongates or prolongs flood events.

The analysis from the technical assessment showed that with option five and six, the duration of Lake Eildon outflows above the minor, moderate or major flood class levels [at Lake Eildon], were increased as a result of these options. They also potentially lead to dam safety risks, because of the increased likelihood of the dam overtopping if the storage was initially surcharged in the first peak of the flood event and then another flood occurs after that. Where the storage is already surcharged, this can lead to potential overtopping risks occurring at Lake Eildon. Options five and six were found to not be a robust way to increase the flood mitigation provided by Lake Eildon.

The two options that did provide flood mitigation benefits from Lake Eildon were option one, which was changing the target filling curve and option two, which was reducing the target storage. However, the cost of offsetting the supply reliability impacts were greater than the avoided flood damages. I'll go into more detail in the subsequent slides about why that is the case.

The influence of the downstream tributaries plays a significant role in understanding what the flood mitigation benefits provided by option one and two, and as they progress further and further downstream at Molesworth and Seymour. On this slide here, I've presented the options which look to reduce the target storage to 78 percent, 85 percent, 90 percent, 95 percent of full supply level and there are the flood frequency curves plotted at Eildon, Molesworth, and Seymour. On the Y axis, is the flow that's presented in cubic metres per second and then the different flood class levels at Eildon and Seymour are presented with a dashed horizontal line, which denote the flows in megalitres per day.

What the key messages out of these plots is indicated is that at Lake Eildon there are some flood mitigation benefits provided by these options. However, as you progress further and further downstream to Molesworth and to Seymour, the lines or the degree of separation between the lines decrease as you move downstream. This indicates that under different peak flow conditions and with the different operating options in place, that they would result in similar peak flows by the time you get to Molesworth and even more so at Seymour. This indicates that the flood mitigation benefits decrease, shown as the lines get closer and closer together.

I haven’t shown all the options to change the target filling curves, however they are presented within the technical assessment report. But they were also similarly influenced by the impact of the downstream tributary inflows. The tributaries that influence the flows at Molesworth include the Rubicon and the Acheron rivers and then the tributaries that influence the flows at Seymour include the King Parrot, the Yea, the Sugarloaf and the Sunday [Creek] river systems.

As part of the water resource implications modelling, the volume of water required to reinstate the reliability back to the base case condition was explored as part of the technical assessment. The volume of the water to offset the changes to the low reliability water shares was estimated by matching the base case volumes. For the table that summarised here, in the first column the volumes are presented and then in the second column is the approximated initial capital costs. To estimate the initial total cost of returning the reliability of the supply back to the base case, the recent trading price of low reliability water shares, which was $1,000 per megalitre was used to multiply by the volume to offset the changes to approximate the initial costs. For example, for the 75POE [probability of exceedance] target filling curves, the volumes to offset it was up to 10,000 megalitres, which equated to $10 million of initial capital costs for the reduced target storage option. For the 95 percent target storage, that was estimated to be $20 million, for the 90 percent was $50 million and then for the 85 percent and the 78 percent, they were $100 million and $155 million.

A range of different cost estimates has been provided for the 85 percent and the 78 percent target storage options. That's because there wasn't necessarily one volume that was able to match the base case reliability and therefore a range of different volumes were trialled. Therefore, the costs for the 85 percent target storage could be as low as $60 million or could be as high as $266 million. Similarly, with the 78 percent target storage option it could be as low as $80 million, and it could be up to $266 million.

Something to note with these initial capital costs that are presented here on this slide, is that the costs don't include the foregone production [impacts] if the volume of water available for consumptive use in the Goulburn system is reduced. The cost also doesn’t include the modification of community assets, such as boat ramps, within Lake Eildon. They also don't include the cost of offsetting the reduced income to GMW [Goulburn-Murray Water] from fees associated with storing water if entitlements are retired from the Goulburn system. And so, we recognise that these are some preliminary cost estimates and there are also some costs that haven't been accounted for.

In terms of looking at the other side of the benefit to the cost ratio. The tangible flood damages have been estimated using a Goulburn Broken Catchment Management Authority hydraulic model of the Goulburn River system. This was used to represent or approximate the inundation extents for various different flow rates along the Goulburn River system. It's important to note with the assessment of the tangible flood damages that they've been assessed between Lake Eildon and Seymour, and so the flood damages haven’t been assessed downstream of Seymour. They also don't include the flood damages along some of the tributaries that feed into the Goulburn River.

The costs have been calculated for residential buildings, non-residential buildings, such as commercial and industrial buildings, road damage and replacement, agriculture, as well as indirect costs. This includes temporary accommodation costs or the cost of emergency management. It's important to note with the agricultural damages that they have been calculated based on a peak flow rather than [including] the consideration of the duration of inundation. So, whilst recognised as part of this assessment, they could be underrepresented or underestimated because they don't include the duration of the inundation. There are also other costs that haven't been factored into the tangible flood damages such as the mental health, the impact on the social welfare as well as the environmental recovery costs.

So, on this slide here, presenting the flood damages estimated for the 1993 flood event along with the 2022 flood event. And on here, there's the base case, which represented an estimate of the flood damages during those flood events. Also included is if the 78 percent target storage option was in place and what impact that could have on the tangible flood damages. The other options of the target filling curves, as well as the other target storage options of 85 percent, 90 percent and 95 percent haven't been plotted here. That's because under those options, if they were to be implemented, the starting storage level within Lake Eildon prior to the flood event, wouldn't have resulted in different outflows from Lake Eildon. Therefore, the cost estimates were similar to the base case estimates presented on this slide here. And just for comparison purposes only the 78 percent target storage has been shown.

Just in terms of speaking for the 1993 event, 50 percent of the outflows from Lake Eildon contributed to the flooding at Seymour, whilst the other 50 percent was contributed by the downstream tributaries. So, the influence or the impact of the 78 percent target storage on the Lake Eildon outflows showed some substantial reductions in tangible flood damages associated, if 78 percent target storage option was implemented.

On the flip side, with the 2022 flood event, the 78 percent target storage would not have significantly changed the tangible flood damages estimated as if that option was in place. That's because of the downstream flow constraints limiting the ability to be able to significantly reduce the peak outflows from Lake Eildon. The total 2022 base case tangible flood damages, as part of this assessment, was estimated to be $410 million. It's useful to compare this with the Deloitte estimate, which was prepared as part of the parliamentary inquiry into the Victorian 2022 floods, which estimated a cost of $432 million, which covered the local government areas of Mitchell, Moira, Murrindindi and Strathbogie. Of those four different local government areas, only the Mitchell and the Murrindindi areas were relevant to the same area of this assessment. Therefore, it's reassuring that the cost estimates produced for the 2022 flood event, and for the base case scenario for this assessment, compared with Deloitte estimate is within the same order of magnitude.

On this slide I’m presented a summary of the tangible flood damages for the different options of changing the target filling curves and reducing the target storage. The avoided average annual damages, or the reduction in in those, were fairly small for the target filling curve options. Where they were approximately $0.2 million, which was around $200,000 on an annual basis. As for the reduced target storage options, they ranged between $200,000 and $400,000 on an annual basis.

So, in order to carry these estimates over into the estimate of the benefit-cost ratio, these estimates have been projected over a horizon of 50 years with a discount ratio of 6 percent and they're presented those on this slide here. In this table, there's two sets of estimates, the GSM estimates and the SGEFM estimates. The GSM estimate is a water resource model that was provided by DEECA and the SGEFM model is a separate model that's been developed by the University of Melbourne. This is referred to as the stochastic Goulburn model, which is being used as a separate model for this assessment to validate the outcomes from the GSM. The SGEFM model has a better representation of storage levels within Lake Eildon in the post-1975 conditions. Therefore, the avoided flood damages and the initial capital costs that have been estimated using that model, are bit different or a bit higher than the GSM estimates. However, the ratios have been calculated from both models do show quite similar ratios.

For the GSM results for the target options, the benefit-cost ratio estimated there was between 0.3 and 0.4 for the different target filling curve options, and for the options to reduce the target storage, they ranged from 0.1 or less than 0.1. For the SGEFM, it also showed for the reduced target storage options a ratio that was 0.1 or less for all, which indicated generally quite low ratios. However, it is also important to recognise that as part of this assessment both the costs and the avoided flood damages do not account for certain aspects. Therefore, if they were to be included, it might not have significantly changed the some of the findings as a result of this investigation. But the ratios do tend to be quite low for options one and two.

But just to provide a bit of a summary for option one and two, these options did provide additional flood mitigation from Lake Eildon. However, the cost of offsetting the supply of the reliability impacts were greater than the avoided flood damages. The low benefit to cost ratios due to the flood mitigation benefits provided were diminished as you moved downstream to Molesworth and to Seymour. This happens because of the tributary flows that happen downstream of Lake Eildon, and as well as the downstream flow constraints limit the ability to maintain that airspace all year round. That's all I had in terms of the presentation from HARC.

Beth Ashworth (DEECA)
Thanks, Hench, for that presentation. Just to recap the next steps, this technical assessment will be available to local councils and Goulburn Broken CMA as an input to the review of flood studies and floodplain management plans. Those studies and plans are the foundation of locally-led and evidence-based solutions in flood preparedness and mitigation.

So that's it for the presentation. We've now got plenty of time to answer questions and I can see a flurry of questions have been coming through throughout the presentation. Thank you for those. If there is anything that we can't answer today, we'll look to publish written responses to those on our website, together with the recording and the transcript of today's webinar.

I'll just look, I'm looking at a different screen to get the questions, so excuse me for a moment. The first question is; why were no agricultural bodies or organisations consulted during the study? While they weren't on the Working Group, AgVic were consulted during the study and provided input on the agricultural damages assessments.

Next question is; where was the data regarding flood frequency at Molesworth taken from, considering there is no working monitoring device available? I'll pass that to HARC to answer.

Simon Lang (HARC)
Yeah, I'm happy to talk to this one Hench. So, flood frequency was estimated at Molesworth using a RORB model, a rainfall runoff model, and that's comprised of many nodes and links. The catchment is broken down into small bits that represent areas that are both gauged and ungauged and then the model gets calibrated at the gauges but can reproduce estimates at other locations. So, ideally there would be a gauge at Molesworth that the model could be calibrated against, it does contain a reporting location there that's been used as an estimate of the flood frequencies, with calibration done upstream and downstream using the available stream flow gauges.

Beth Ashworth (DEECA)
Thanks Simon. And just to add to that, as part of the Victorian Constraints Program, 11 new telemetry and river rainfall sites are being installed, and the gauge at Molesworth is one of those locations. It has required several different approvals which have slowed things down, but those approvals have now been received and construction on that is due to begin in the next week or two. So, I know that'll be good news for some of you for which this has been a priority.

Moving on to the next question; obviously the downstream flow constraints need to be considered in future modelling and better data recording equipment installed. Is there plans for that? And yes, we agree that is a priority and that's something that flood studies led by local councils would identify and prioritise as part of that.

Next question is about; what about a trial period where some of these options are instigated on a small-scale basis to assess their impacts in real time. Surely this could be implemented without a major impact to water supply reliability? The modelling really looks to assess the benefits and impacts of all of the options given the size of Lake Eildon. What the assessment’s shown about how much you would need to hold the lake at a lower level for example, for it to be effective in flood mitigation. There really isn't a small scale or trial option available for that, unfortunately.

Next question is, do you have a graph of the influence of downstream tributary inflows for data for just the past 10 years? Simon or Hench, can I get you to answer that?

Simon Lang (HARC)
So, I'm not sure if this is what their question is getting at, but we did as Hench explained, the assessment using two resource models, one being that GSM from DEECA and that we felt probably underestimated to a degree the observed lake levels over the last few years. And I guess that lake level reflects everything that's going on in the system including downstream tributary flows. So, we repeated the assessment with a Melbourne University model that had a high predicted water level and over the last 10 years and that's why those two sets of ratios were presented at the end. So different models gave different answers, but the ratios between the avoided flood damage and the initial capital cost of the water recovery came out quite similar.

Beth Ashworth (DEECA)
Thanks, Simon. Hopefully that answered the gist of the question. Next question; there was mention of a $10 million budget for flood studies. Will there be an allocation from government to ensure bridges and other assets affected by surcharges can be made more resilient given there's no support for a change in the flood mitigation policy? I think an important part of flood studies is looking at local infrastructure and local solutions for resilience for those assets. I encourage people to get involved, contact their local council about flood studies to ensure that the assets of importance to them are considered.

Next we have; had the January 2024 storm event been experienced 30 kilometres east into the Lake Eildon catchment, it's been suggested that Molesworth, Yea and Seymour would have been terminally impacted. There was reference to earlier flood experiences, 1975 and 2022 being modelled against the options was consideration given to modelling the January 2024 Storm event? So, I’ll pass that over to HARC.

Simon Lang (HARC)
Yes, thanks Beth. The 2024 event itself hasn't been modelled in the same way that the 1975 or 2022 has. But one of the appendices in the report includes recorded stream flow data from that event. That hasn't looked at how the options may or may not have changed what was observed. I guess what I'll say about that event is yes, if it had fallen in the Lake Eildon catchment instead of downstream of Lake Eildon, then outflows from Lake Eildon would have been higher. The impact, I guess, immediately downstream of the dam wall would have been higher. I don't think you can go as far as saying it would have been worse for Seymour, et cetera, given that the Eildon structure itself would reduce the flood peak as it passed through, even if the dam’s full. The dam structure in and of itself does mitigate flood peaks even if the lake is at 100 percent FSL. So yes, if the rain had fallen further to the east, the flood damages would have been different in different locations, but I don't think it would have been worse at Seymour.

Beth Ashworth (DEECA)
Thanks, Simon. Next question is; what comparison has been made between the cost of flood mitigation and the cost of flood recovery? Can I get you to respond to that one, Simon?

Simon Lang (HARC)
Yes. The comparison is between how much it would cost to offset the reliability of supply impacts to the low reliability water shareholders if the lake was to be held at a lower target storage or if the target filling curve was changed. That's on one side, and then on the other side is given that or what's the expected change in flood frequencies and hence flood damages. So that's the two sides of the equation.

Beth Ashworth (DEECA)
Thanks Simon. Another question; can you provide a similar table of supply reliability impacts summary based on the current cost of water shares per megalitre?

Simon Lang (HARC)
Yes, maybe if I can expand a bit out on that question. The cost has been estimated by multiplying the volume of low reliability water shares that would need to be recovered or retired so that water share holders aren’t impacted. That has been multiplied by recent cost of the permanent transfer of low reliability water shares on the water market, which is $1,000 per megalitre. If you want to estimate other costs for regaining that water or recovering that water, simply multiply the volumes in the report by the dollar per megalitre value you chose. The reason to use the permanent trade of $1,000 megalitre option is to offset those lower reliability water share impacts from having to lake at a lower level all year round. We think it would be important to provide the additional flood mitigation you would need to recover those water shares permanently. Hence why that $1,000 per megalitre value being used rather than allocation trade value in terms of dollars per megalitre.

Beth Ashworth (DEECA)
Yeah, and just to add to that, there has been some discussion about whether you could use temporary trade as an option to achieve the same outcome. And you know, I think to maintain the desired amount of airspace, you'd need to be continuously buying and releasing allocations to match the inflows so that you weren't impacting on other people's reliability. Which would in itself, the constant buying and releasing, would have potentially significant market impacts. This may potentially drive up the market price for other participants and so on. As the report points out, you might not be able to release all of that water as well because of the downstream flow constraints and the need to prevent flooding impacts on downstream communities.

Moving on to the next question; Will DEECA be recommending that some funds obtained from water entitlement holders be quarantined to support rebuilding damaged private and public assets from storm events, which are occurring increasingly more often? So no, that's not something that we'll be recommending. Just to reiterate a couple of things that we spoke to earlier. This technical assessment has found that the flows from downstream tributaries, rather than releases from the lakes have more significant impact on flooding in townships like Seymour and Molesworth. Also, I just want to point out that Lake Eildon does provide significant flood mitigation benefits already. As I said, in the October 2022, the existence of Lake Eildon reduced the peak flow by 70 percent. So just wanting to reiterate a couple of those points.

Next question; will the recommendation to the minister include the economic, financial and social impacts of the floods rather than just the value of water for irrigation in a potential drought? If not, when would a broader reality of flood impacts become contemplated? I think I'm keenly aware, after many discussions that I've had with community members, about the really traumatic impacts, - [not just] financial, but physical, emotional, mental impacts - of the flood events and I really believe you can't put a cost on that. So, I just really want to recognise those impacts. Even though the technical assessment doesn't put a dollar value on them, I want to acknowledge that those impacts are there. The investment principles of the Floodplain Management Strategy really indicate that to invest in flood mitigation benefits, the benefits in terms of avoided flood damages need to outweigh the capital and ongoing costs of those. We wish this technical assessment had found an option that warranted further investigation in that respect. But unfortunately, the evidence isn't there to support that.

Next question; where was the data on agricultural impact gleaned from for this study? Simon, can I get you to respond to that one?

Simon Lang (HARC)
That's outlined in one of the appendices to the technical assessment report.  It includes a description of the method used to estimate the flood damages, so feel free to please check that out. At a high level, it's a rapid assessment method developed in the early 2000s, which looks at unit costs for various land uses. The hydraulic model was used to estimate the area of land inundated at various flow thresholds, and then how that changed according to options. And then, those areas inundated were multiplied by the unit costs, after adjusting them for inflation since they were set. Have a look in the appendix for the various unit costs used.

Beth Ashworth (DEECA)
OK next question, the duration of inundation has not been considered. Why has that not been taken into account?

Simon Lang (HARC)
One of our, I guess challenges, with the technical assessment was only being able to use models currently available. For the water resources component that was a monthly time step model. Ideally if we'd been able to use a daily time step model, which hopefully be available in the near to mid future, then we could have accounted for duration impacts. At the moment, we've got a monthly water resource model, complemented by an hourly time-step rainfall-runoff or flood models. So, looking at those duration impacts of whether an option reduces flooding from a couple of weeks to a few days wasn't possible with the tools we had to hand.

Beth Ashworth (DEECA)
And just to add to that, the use of a daily time step model would have delayed the assessment by a number of months, which we know people were very keen to get these results. The use of the monthly time step model, as I understand it, and Simon you can confirm this, was considered fit for purpose for this assessment. The use of a daily time step model, given the orders of magnitude difference that would have been required in the cost benefit analysis, isn't expected to have impacted the findings of the assessment. Can you just confirm that?

Simon Lang (HARC)
Yes. The assessment and the way it has been done and written up, certainly highlighted aspects that could have been done better. But we don't expect that doing them to a higher level of detail will have changed the findings. The damages would need to change by, you know, orders of magnitude for the ratios to get closer to one to one.

Beth Ashworth (DEECA)
Also, that 30 percent uplift that's been included in the avoided damages estimates, would go some way to account for, for that duration impact as well.

I think that's all of the Q&A's that we've got, but let me just confirm that. OK. I think that's all we've got for today, so that brings the webinar to a close.

We hope it's been useful for you in understanding both how the assessment has been done and the findings of it.

Thank you all for making the time to participate today. Really appreciate it and for all your constructive and useful questions.

Thank you very much. Take care all.