Wivenhoe pumped hydro: the big little plant that didn’t

Queenslanders would pay less for power if the generation sector was split up

Wivenhoe power station. Source: Queensland Mining & Energy Bulletin

The Wivenhoe pumped hydro power station in Queensland should be making huge profits. By our analysis, if it had run flat out in the late afternoon window this summer – from 4.30pm to 7pm – it would have made profits of more than $50 million.

But it didn’t actually run that hard at all. At no stage did it produce more than 50% of capacity and it only ran on 59 – or about one quarter – of the 250 highest price half hours so far in Queensland this year. And it is the Queensland consumer that has suffered the most.

Why is this? Basically, it’s because the vast majority of generation in Queensland is owned by two companies -CS Energy and Stanwell. CS Energy, the owner of Wivenhoe, also owns significant coal generation.

It’s much cheaper to run the coal generation instead of Wivenhoe and so, for the most part, Wivenhoe doesn’t run. Of course, CS Energy could run both Wivenhoe AND its coal generation, but if it did that it would depress prices in the afternoon peak, negatively impacting its coal profits more than the positive impact of the extra capacity.

As a result, Queenslanders pay more for their power than they might and this summer have been paying more than 50 per cent more than other states, including and most particularly renewable-rich South Australia. The high pool prices although mostly hedged in the first instance, but they will eventually find their way into consumer prices.

CS Energy, from its own point of view, is doing nothing more than profit maximising.  This is legal and there is, in our opinion  nothing wrong with that, it’s basically the objective of business to maximize the average return on capital employed.

However, we are (very) critical of the generation ownership arrangements in QLD that make it “sensible” not to operate peak generation when prices are high. We drew attention to  the gas oligopoly in South Australia but the coal oligopoly in Queensland costs consumers far more.

It’s our strong view that the generation industry in QLD is too concentrated and  that  Wivenhoe should not be owned by a coal generator that has more to lose than gain by running it.

One of the many advantages of renewable energy in general is the number of suppliers.  It’s why household PV and household batteries have advantages to consumers way beyond those that are obvious in the initial price.

The collective impact distributed generation has is enormous and the good thing is most consumers aren’t even aware of their influence on the broader market.

About Wivenhoe and QLD generation

Wivenhoe pumped storage consists of 2 x 250 MW units in QLD.  A report on the CS Energy web site (owner of Wivenhoe) states that in September, 2016 a $31 million refurbishment of one unit was undertaken and was completed in November 2016.

CS Energy is one of two QLD Govt owned generators in the state. Besides Wivenhoe it also owns the Kogan Creek station, Callide A & B, a 50% interest in Callide C and the Gladstone power station. The website states that CS energy can trade 4,035 MW in the NEM.

The other QLD Govt. owned generator is Stanwell Group which can trade 4,100 MW of power

Average demand in in Qld for the 12 months ended 31 December 2016 was 6000 MW, so between CS Energy and Stanwell it could be argued the QLD Govt has a strong grip on the market.

It may not matter that both groups are owned by the Government because with just two dominant generators the tendency for an oligopoly to emerge is inevitable.

The fact that they are Govt owned simply means that the QLD Govt doesn’t have much incentive to change it.  Indeed the QLD Govt was responsible for the oligopoly since it converted three companies into two a couple of years  back.

We’ve previously pointed out that the way the NSW QLD transmission is designed, its much easier for power to get out of QLD than it is for NSW Generators to get their power in.

In addition, over the past two-three  years there have been various “raised eyebrows” to put it politely in the way the QLD generators have used the 30 minute settlement rule in QLD.

At one stage this was occurring, in layman’s terms, by very short term transmission constraints in North Queensland. As I understand it, this means that North Queensland generator (say Gladstone power station) sets the price North of the constraint and this price then becomes the Qld regional node price  for the five minute interval.

Southern Qld generators just North of the Qld border hog the transmission into South East Qld and so NSW Generators haven’t got much of the action.

If the Federal Government wanted to make a start on effective  energy policy in a non ideological way it could maybe start by getting the  ACCC to look into the QLD arrangements more closely.

In any event that’s by way of background.

Wivenhoe doesn’t run much even when prices are very high

Wivenhoe is a peak generator so we don’t expect it to run very much. We do expect it to run though when the price is high in QLD. Considered in isolation Wivenhoe would/should run when revenue exceeds variable cost.

However mso far this calendar year in QLD despite record demand, maybe 20% above previous records at times Wivenhoe has not operated at more than 50% of capacity at any point.

We took the top 250 half hourly price intervals in Qld this calendar year and looked at Wivenhoe output for each of those times. It only had any output on 59 half hours (20% of the time and only 50% of capacity at best). This is shown in the following figure. All the dots on the X axis are when it didn’t run. Correlation between pool price and Wivenhoe output is non existent. Of course this could be because the water storage was empty, but somehow we don’t think so.

Figure 1 Wivenhoe output at high prices this year. Source NEM Review

Coal generation is more profitable

A far more likely explanation is that there was enough coal and gas power in Qld to cover demand and it was more profitable to run the coal generation than to run Wivenhoe.

CS Energy’s coal plants have short run marginal costs [SRMC] of about $18-$20 MWh. As we show below the SRMC of Wivenhoe is around $100 MWh at the moment. Provided the lights aren’t going to go out CS Energy makes more money running coal than Wivenhoe.

That’s just from a cost perspective.

It could run both Wivenhoe and the coal generation but that would lead to a lower pool price and since CS Energy has 5-7X more coal capacity than pumped hydro it would more or less  lose $5-$7 on coal for every $1 of Wivenhoe gross profit. So its perfectly rational from both a cost and revenue point of view for CS Energy not to run Wivenhoe much.

Wivenhoe’s SRMC around $100 MWh

The SRMC of Wivenhoe is basically the pool price of electricity when the pumping takes place, times the inverse of the efficiency factor.

In calendar 2016 Wivenhoe produced 114 GWh  and used 172 GWh in pumping. That’s an efficiency ratio of 66%. As an aside because the pumping comes from coal powered generation its actually carbon inefficient to run. CS Energy also states its liable for REC and SREC surrender when pumping (consuming electricity).

So if we look at our increasingly familiar figure of the QLD pool price by time of day so far in 2017 it looks as follows.

Figure 2 Pool prices by time of day 2017 Source: NEM Review

Although its not clear from the chart in the 1:30 AM to 6:30 AM window the average pool price is around $61 MWh. If we allow for the efficiency factor that might come to around $90 MWh if we add in the present LRET and SREC liability that might take it to $100 MWh

So that’s about 5X the $18 -$20  cost of coal generation and the Wivenhoe variable cost is comparable to the variable cost of  open cycle gas at least during this Summer.

Also if you increased the demand in QLD by 250 – 500 MW during that time period it will surely increase the price. We don’t have enough data to say how much. We could maybe work it out but too much trouble, probably no more than $20 MWh and a guess would be $10 MWh

Equally dispatching 500 MW of additional power will lower prices, again we don’t know how much and its much harder to work out.

Still Fig 2 indicates its fairly profitable to operate in the 4:30 pm to 7:00 PM window.

As it happens the median (half prices above, half prices below) price this year in QLD in that window is $145 MWh but because of some high price events the average price is around $838. Some statistics that dimension prices over that time frame this year are shown in Fig 3.

Figure 3 Pool prices in the afternoon peak QLD 2017. Source: NEM Review

Our conclusion is that it would have been profitable 75% of the time to run Wivenhoe flat out in the afternoon peak.

If we use the average price of $838/MWh, and assume a variable cost of $100, then we estimate that if Wivenhoe was a merchant generator, and before allowing for the difficult to estimate impact its dispatch would have had on price, then it would have made a gross profit of around $65 m so far this year by dispatching in the afternoon peak and pumping in the wee hours of the morning.

Figure 4 Wivenhoe potential profit. Source ITK

---

David Leitch is principal of ITK. He was formerly a Utility Analyst for leading investment banks over the past 30 years. The views expressed are his own. Please note our new section, Energy Markets, which will include analysis from Leitch on the energy markets and broader energy issues. And also note our live generation widget, and the APVI solar contribution.