Solar, wind farms hit as AEMO slashes output calculations

Broken Hill solar farm
Broken Hill solar farm

New calculations on line losses and grid congestion by the Australian Energy Market Operator has resulted in some new solar and wind farms suffering major losses in their calculated output – up to  22 per cent in the most extreme case.

The worst affected by AEMO’s newly released “marginal loss factor” calculations are those wind and solar farms located furthest from the main load hubs, in north Queensland, in western NSW and some in Victoria. Many projects have suffered cuts of between 10 and 22 per cent.

The MLFs are a critical part of a business case for any sort of generator, renewable or fossil fuel. They act as a “multiplier” of revenue and big cuts can seriously affect the business plan of a new or existing plant, and could “make or break” a project.

The MLF calculates the difference between how much is produced by the power facility, and measured at its meter, and how much is estimated to be delivered to customers, and so how much is paid, or credited, by AEMO.

They also apply to consumers, meaning that a manufacturing plant, for instance, located a large distance from supply will have to pay a “multiplier” of its actual demand to account for the anticipated losses over the network.

The calculation depends on a range of factors – the quality and length of the line, the existence or distance of local demand, and how much other generation is in the same area. And the estimates change each year.

The changes unveiled by AEMO in late March are dramatic for some plants, and represent what the market operator says is a major change in the way electricity is flowing across the grid.

In Queensland, for instance, the electricity is “flowing south”, rather than north, because of the increase in generation in the northern part of the state and the reduction in load in central Queensland.

However, the new assessments are being challenged by some wind and solar developers, and has sparked debate about the methodology used and the variability from year to year.

“These MLFs are essentially a multiplier for revenue of a plant and affects the economics in a material way,” said one project developer. “It’s going to be a real issue.”

A spokesman said AEMO has held discussions with stakeholders about the need for change in 2018/19 and will continue to listen to any stakeholder feedback.

“We are currently looking to consult with the market participants regarding possible changes to these MLF processes,” he said in an emailed statement.

“Initiatives such as the Integrated System Plan, looking specifically at renewable energy zones, and holistic national planning aims to optimise the power system, putting downward pressure on MLFs in the future.”

The worst hit in the latest assessment is the 53W Broken Hill solar farm, owned by the AGL-linked PARF, which has had its MLF calculation slashed from 1.2456 to 0.9789, or 22 per cent.

That means that for every 100MWh it produces, it will get paid for 97.89MWh. It had been receiving 124.56MWh.

(Ironically, the previous estimate of 1.2456 meant that it was deemed to be improving and benefiting the grid, before the new assessment. Many wind and solar farms still have an MLF of more than 1.0, meaning those facilities are considered a net benefit to the local grid).

In Queensland, the worst affected are the newly completed 50MW Kidston solar farm, owned by the listed company Genex, which is trimmed from 1.0115 to 0.8979, a loss of more than 12 per cent.

The soon-to-be-completed Clare solar farm suffers a similar percentage cut, going from 0.9823 to 0.8727, as does the Hughenden solar farm, also near completion, which is trimmed from 1.0115 to 0.8979.

The smaller Longreach solar farm goes from 0.9689 to 0.8934, and the Barcaldine solar farm suffers a similar reduction.

In Victoria, the Ararat and Kiata wind farms are the worst affected, suffering losses of 5 per cent, while the newly connected Gannawarra solar farm is also hit with a reduction to 0.9729 from 1.044 – a rating that will apply even though it will have a battery added by the coming summer.

In South Australia, there is little change, although the new Tesla big battery at Hornsdale has had its MLF reduced slightly on its generation side –  to 0.9771 from 0.9886 – along with a smaller reduction on its load side.

Wind and solar projects are not the only ones affected. Major coal and gas fired generators have also had their MLFs reassessed but the changes for these mostly centralised plants are mostly minor – some are improved, while many are also reduced slightly.

However, regional diesel generators and other peaking plant located at the edges of the grid are also badly affected for the same reasons as wind and solar farms.

These include Origin Energy’s Mt Stuart peaking generator in north Queensland, which runs on jet fuel, and is judged to suffer similar losses to solar farms in the area such as Kidston, Clare and Hughenden.

AGL’s Wattle Point wind farm in South Australia – soon to have a battery storage facility added – has the worst MLF of 82.79, probably because of its location at the end of a long line on the Yorke Peninsula.

The 91MW wind farm is only credited for 82 megawatt hours out of every 100 megawatt hours it produces. That means developers of such projects have to make sure the wind resource is good, or the price they are receiving is better than elsewhere.

Goldwind’s White Rock wind farm in northern NSW has a very low MLF of 0.8413 – also  little changed from this year.

RenewEconomy talked to number of project developers and owners, who preferred not to be named because of the sensitivity of the issue.

They said there are major issues about the MLF calculations, which might have made sense in a centralised grid, but may need to be reassessed in a modern grid with distributed resources.

They noted that AEMO’s push to create renewable energy zones to group wind and solar projects in the same area may prove problematic, because it would risk similar assessment of over-production in one area, and cause a similar cut to assumed MLFs.

AEMO’s explanation centred on the increase in renewables in some regions, the decrease in loads elsewhere, the change in the direction of flows to and from a regional node, and the change in patterns of use in interconnectors linking different states.

This graph to the right illustrates the power plants that suffered decreases in Queensland (purple) are all located to the north. Some in the south-west had their MLF increased due to lower fossil fuel generation.

AEMO noted that MLFs for connection points in Broken Hill have decreased by up to 22.7 per cent, including the solar farm.

“New generation projects connected near Broken Hill have resulted in projected generation in the area increasing by over 85 per cent, increasing the projected power flow toward the RRN substantially,” it noted.

That suggests that the Broken Hill solar farm is affected by the construction of the new 200MW Silverton wind farm, also owned by PARF.  MLFs  at the connection points close to Parkes, Forbes, and Griffiths – where three new solar farms have recently been installed – had also declined by up to 5.6 per cent.

“These are risks that are only just starting to be understood,” said another developer. “It does create a new problem.

“There is talk of some review. The MLF calculations were set up in a world without distributed generation. Are they the right multiplier in a worldd of distributed generation.”

Another noted that the wind and solar farms in north Queensland had been worst affected, and noted a similar assessment would be made of any new coal generator in the region, although it is highly unlikely any such development would take place.

Update: An AGL spokesperson said in an email: “We’re aware of the changes but have no comment on the impact and plan no action.”

 

 

 

 

 

Comments

38 responses to “Solar, wind farms hit as AEMO slashes output calculations”

  1. Chris Fraser Avatar
    Chris Fraser

    Does this put pressure on locations for coal generators, which are so toxic they have to constructed a long way from everybody ?

    1. Joe Avatar
      Joe

      Public Health Alert – All Coalers to be Shutdown Forthwith. Oh wait, that can’t happen. Public Health always ranks last.

    2. Giles Avatar

      Yep, it applies to all. Just it hits wind and solar farms the most cos they built further away. usually.

      1. John Saint-Smith Avatar
        John Saint-Smith

        ‘Further away” is where they are in relation to demand at the moment, because relatively little manufacturing and material processing occurs in the more remote, ‘solar friendly’ parts of the grid. But surely that could change as the concentration of renewable energy generators in those regions grows and attracts more users?

  2. Allan Barr Avatar
    Allan Barr

    There is no way transmission losses are 22%, is there?

    1. Giles Avatar

      read again – the change is 22%, because previously it got a significant credit, because it was pumping into a local grid when previously the locals had to import from a long way. but the addition of silverwater wind farm means it will all have to travel the other way. wattle point has worst MLF, 82. something, which means line losses of nearly 18%, which pretty hefty i have to say.

      1. Jon Avatar
        Jon

        Silverton Wind Farm

  3. Cooma Doug Avatar
    Cooma Doug

    This is day one of a long process of adaptation of many different products. This might sound odd but nobody knows yet where it is going to be in 10 years.
    An example:…Fred who lives in Victoria will be able to be in a load shifting and sharing arrangement with Bill who lives in Brisbane. Neither of them will actually physically use or share from each other. They will be using virtual power.
    The market will scrutunise all bids as now, but there will be many more factors.
    This might seem odd when Bill is functioning on his own battery for a time and fred is feeding his next door neighbour. The market crunches the numbers to create the virtual settlement. These choices are done to optimise transmission effectiveness and security.
    The reality is that grid solar and wind as well as batteries and roof top solar will have much less transmission costs with load side market functions gaining ground.

    Sorry if it sounds like disney land stuff. But if you spend a day with energy traders you will walk away thinking it is disney land.

    1. Steve Jordan Avatar
      Steve Jordan

      Glad to see you commenting on this, Doug. And, Giles, could we get some comment from the other people in your stable of well-informed commentators?

  4. bedlam bay Avatar
    bedlam bay

    Surely these remote generators would have factored in these items. Or has Team Turnbull changed the goal posts (like Abbott who sent out a very embarrassed Ian McFarlane to hit the renewables sector). And has Turnbull and Coalavan leaned on Alinta to make an offer for Liddell.

  5. Robert Johnson Avatar
    Robert Johnson

    Bit rich of inexperienced developers and investors to be seeking a change in the rules because they didn’t (probably still don’t) understand MLF. All developers complaining about MLF changes indicates is that sufficient understanding of MLF didn’t exist in those developers and investors and their advisors.
    MLF and is not new.

    To change the calculation methodology would only serve to bail out those who made poor decisions in the first place. Who will pay for the bailout? Consumers?

    By and large the logic of MLF is sound – you do need to account for losses in the network somehow. The alternative to the current methodology of applying a loss factor to generators is to put those losses onto consumers – the downside of which would be a complete absence of a price signal to build generators in efficient parts of the network where losses are lower and the actual energy generated makes it to consumers. Hence the overall cost would go up.
    Normally the price signal (MLF) is sufficient to mute the impact of MLF on projects. One commits (or gets built) and the competent developer nearby recognizes the cumulative impact of their project and decides not to proceed – or maybe their project is profitable enough that it can proceed anyway and the first project owner also had a bit of buffer in the economics and can also sustain a reduction in their MLF. Unfortunately, what we have seen in the past few years is a lot of highly aggressive developers who don’t understand the network (or who are working on the principal that there is a greater fool to flip to) riding the high times in the market with little room to move in their project financials going headlong into projects.
    The impact is also mitigated through site selection. Put a generator in a strong part of the grid or near a reliable load and the MLF is much more stable. Its common sense really.
    Much of the current dynamic has been driven by a drought of offtake a few years back and then a sudden flood of offtake contracts (so retailers can meet their RET obligations). Had projects been committing progressively then the changes would have been progressive rather than sharp and even the less competent developers and investors would have paid closer attention to this issue and the impact on their projects. This doesn’t excuse a lack of understanding of the MLF methodology and likely impact on any given project though.
    Notably the impact is greatest where generators have similar generation profiles such as the zones being proposed, clusters of solar together means lower MLF for the cluster because they are all generating at the same time and heavily loading the network, thus increasing the losses. This is overcome by strengthening the grid to the zone which reduces the losses, but someone has to pay for the grid augmentation too….
    Interestingly, this is not the end of the MLF falls, there are still projects being built in some of the more impacted parts of the grid which will drive the MLF’s lower in those locations.
    Stay tuned for the next exciting episode …. network constraint impacts.

    1. Guy Perkins Avatar
      Guy Perkins

      System may be imperfect but agree that developers and owners should be able to understand the rules of the game and site projects is less risky areas, or price the risk in to begin with if not. Worried new investors have blindly followed forecasts that are notoriously dependant on a suite of input assumptions that are totally expected to change over time.

  6. Jon Avatar
    Jon

    These numbers will be very dynamic over the transition period we are in.
    Every new decent size generator or load added or subtracted will shift the balance and flows in the grid.

  7. DJR96 Avatar
    DJR96

    OK, can someone confirm my thoughts here.
    Take the Broken Hill solar farm for example. When it is generating during the day, it may be supplying enough to power the whole town and region and some excess gets pushed back to the wider grid. So surely the MLF should only be applied to the portion of energy going back to the wider network, not all the energy being used locally.

    Is that the way MLF is calculated? Or have they completely missed this essential point.

    1. Jonathan Prendergast Avatar
      Jonathan Prendergast

      It sounds to me like you are on the right track. If MLFs were high before, this suggests losses to get electricity to Broken Hill customers, so they pay a premium. Now the solar farm has reversed this, and there are more losses moving the solar back up the grid eastwards. I believe it is averaged of losses of all generation and supply.

      Maybe you are onto something.

      1. DJR96 Avatar
        DJR96

        Setting a fixed/constant MLF doesn’t sit well with me. It’s not right.
        The power flows are all quite dynamic and can even swing back and forth. There is enough metering points throughout the network to be able to dynamically work out where the losses are and therefore who to attribute them too.
        The existing arrangement is clearly designed for the old conventional network, and is not suitable any longer. Trying to adapt it is not working well and some entities are getting shafted at least some of the time.

        1. Jonathan Prendergast Avatar
          Jonathan Prendergast

          In this case, Broken Hill Solar farm are cross-subsidising Broken Hill electricity customers. Maybe solar farms should be smaller?

          1. Malcolm M Avatar
            Malcolm M

            Rather the opposite, that Broken Hill consumers were paying 28% more than the NSW reference price, supposedly because of line losses. This recent analysis shows that output of the solar farm was mostly consumed locally, so line losses were minimal and there was no need for the premium.

        2. Malcolm M Avatar
          Malcolm M

          For market simplicity a constant loss factor is used within regions but a dynamic loss factor between regions. (The dynamic loss factor varies with the load.) If there are parts of the intra-regional network where constant loss factors are inappropriate, a case would need to be put for a rule change. And we know how hard that can be.

          http://aemo.com.au/-/media/Files/Electricity/NEM/Security_and_Reliability/Loss_Factors_and_Regional_Boundaries/2018/Marginal-Loss-Factors-for-the-2018-19-Financial-Year.pdf

    2. Giles Avatar

      My understanding is that solar farm got a high multiplier cos only local generation, therefore avoiding line losses from electricity coming from lithgow coal, for instance.
      With Silverton, there is excess capacity, so most it will be all travelling back t’other way. so silver ton and solar farm now have similar ratings (less than one) that shows losses going in other direction.

      1. Robert Johnson Avatar
        Robert Johnson

        you understand this stuff Giles, don’t be surprised if you get a few job offers from Developers 😉

        1. Giles Avatar

          Gee, if they haven’t got that far then they shouldn’t be in business. but,some are probably right to be surprised by scale of change. Fair point someone made here that in a distributed world that even aemo say we’re heading to, this shouldn’t happen. just goes to highlight the scale of change needed at so many levels.

      2. Malcolm M Avatar
        Malcolm M

        The report shows a multiplier of 1.0062 for the Silverton Wind farm. However Silverton was not listed as a new generation unit considered in the simulations. The 2019/20 marginal loss factors for Broken Hill could be quite different again, because there will be both more generation (from Silverton), but also more load. There is a magnesite mine being built to the south-west of Broken Hill, which will have an average load of 87 MW, and the North Broken Hill mine will be reopened. The 200 MW from Silverton at a 50% capacity factor, coinciding with ~100 MW of new demand, should approximately balance, leading to a multiplier close to 1.0. The beneficiaries of this local generation would be the mining companies.

    3. Robert Johnson Avatar
      Robert Johnson

      This is correct. Its based on load flows in the path(s) to the node. So if the generator profile perfectly matched the load or never exceeded it then the MLF would be 1.00

    4. solarguy Avatar
      solarguy

      I was thinking along those same lines as you. But just as I thought I was starting to understand it, further reading just confused the shit out of me. The examples of the % losses don’t compute.

  8. john Avatar
    john

    As I see it if a Coal Generator is working then any RE that is supplying more that is needed is going to be turned off.
    If the demand is above what the Coal Generator can deliver then solar and wind will be t

    1. Phil NSW Avatar
      Phil NSW

      OMG – Reverse logic at its best. Solar and wind must have guaranteed supply contracts as first preference due to their zero carbon emission. This is bureaucracy gone mad.

      1. hydrophilia Avatar
        hydrophilia

        Hmm… complex possibilities. I suppose one must consider both price and reliability and who will offer you the best deal on the combination you require. In my books that would be solar/wind + storage, but YMMV.

        1. Phil NSW Avatar
          Phil NSW

          The coal fired power station is an adjustable supply generator. Remember how when there was no solar feed in, the power stations maintained the grid in balance. They still can do this. In the future if supply exceeds demand the excess goes to battery/PHES. Zero emission sources have to be first preference in this equation. If the NEG is truely bypartisan this has to be the operating model. Fiddling the edges with the MLF to penalise RE is demonstrating there is insufficient inter-connectability. Our gold plating mates need to address these issues.

          1. Andrew Roydhouse Avatar
            Andrew Roydhouse

            Problem with the coal generators is they cannot increase/decrease output as quickly as it changes with the initial PV ramp up/down.

            Time for some batteries….

  9. Colin Nicholson Avatar
    Colin Nicholson

    I thought decentralisation away from the cities was on the agenda This is hardly helping unless MLF operates nett

  10. HybridPower53 Avatar
    HybridPower53

    Obviously, they would be paying a premium for power coming in to their location so to overcome these inefficiencies they’ll have to install batteries????. BRING IT ON!!!

  11. Jon Avatar
    Jon

    Thanks Giles
    It would be interesting to see modeling of how the MLF will shuffle with Liddel closed.

  12. Chris McGrath Avatar
    Chris McGrath

    MLF is a very important driver to encourage generation and load to be established in appropriate places for the network and this is a good thing as increased renewables uptake occurs. It is (or at least should be) a key factor in the siting and sizing of new projects by developers

    Conducting forecast MLF studies is a basic part of project due diligence. The example of Wattle Point was one project that learnt the hard way before these forecasts were standard practice. The early development of the Moree Solar Farm (developed by BP at the time) was a great example of a project that learnt the hard way by omitting to do a forecast when it was standard practice – late in its development (and only after winning the Solar Flagships and failing to build) – the project was reduced from 150MW, which would have resulted in a (easily forecast) reduction from an MLF of 1.03 to 0.86, to ~50MW, an appropriate size for the network and a project that has probably retained a stable MLF.

    I’d be interested to know whether the example projects in the article with these swings in MLF had conducted MLF forecast studies prior to construction and what these predicted? If they engaged a good consultant to do these, and got their inputs and assumptions correct (this is probably the hardest part), they may well have known this change was coming and proceeded to build projects on the basis of this. I imagine that AGL were well aware of the impact coming for Broken Hill Solar Farm (I also can’t imagine they had banked retaining that 20% uplift from MLF in their long term economic modeling).

    The more important question to me is – if the developer followed best practice in forecasting MLF and it still varied wildly from the best forecast available at construction – why did this occur and, subject to this answer, is this a fair risk allocation as the network changes its behaviour over time?

  13. George Darroch Avatar
    George Darroch

    Although they’re not subject to MLFs, this does demonstrate some of the value of rooftop solar installations, where the line losses (although not system losses) are zero.

  14. Peter F Avatar
    Peter F

    This is one of the better debates I have seen on RE. Whether the actual numbers are right or wrong I don’t know but the policy is correct. Renewables should be built near the load or the load built near the renewables as has normally been done with Aluminium and hydro.
    If renewables are really cheap at Broken Hill maybe next stage processing should be located there helping decentralisation and minimizing transmission and or freight costs

    1. Mike Westerman Avatar
      Mike Westerman

      Transmission is strongly influenced by regional development politics – development of natural resources, attraction of population and industry, etc. This is the case if energy hubs are developed.

  15. Ray Miller Avatar
    Ray Miller

    After reading your article Giles and the current comments it appears the current MLF is pseudo science and very much past it’s used by date.
    Maybe the whole exercise is designed to add more pressure on the AEMC to bring all the NEM rules into the 21st century?

    How can a liquid fuel peaking plant in Townsville have it’s MLF reduced when the very point and location of the plant is to supply loads during scarcity of supply which are most likely closer to the peaking plant then the southern generators?
    As has been pointed out both the time and location dynamics of the loads on the national grid are complex, granted we should be striving for a closer match between local load and generation.
    Which again raises the issue of rooftop solar and battery storage which is behind the meter, arguably is the most advantageous location, with any exports only to the local area with minimal transmission losses. Maybe this idea will catch on? Then the AEMC rules will end up being even more irrelevant to the current situation?

Get up to 3 quotes from pre-vetted solar (and battery) installers.