Dialogue: Why the wholesale power market just cannot work

Soaring wholesale power prices are the inevitable consequence of an electricity market that carries the seeds of its own destruction, writes BRIAN LEYLAND*.

Until recently, the wholesale price of power was 4c to 5c a kilowatt-hour - less than before the market was set up. But low hydro storage levels have seen prices regularly between 15c and 30c a kWh and as high as $1.

Some see this as the market working as it should. Others disagree and have called for regulation.

Our wholesale electricity market is based on an economic model intended to provide the lowest possible wholesale cost of power while providing a reasonable income to the generators.

In similar markets in Australia and the United States, shortages and transmission constraints have driven the cost of power up to several dollars a kWh. The market has failed in California, where price caps are being imposed, and there have been huge price spikes in the north-eastern US.

A paper from the US goes back to the economic model to explain why the market has failed. It concludes that to work, the market must have several features.

These are competing generators; thermal power stations which have a high marginal ( fuel) cost and a relatively low capital cost; more than sufficient plant available to meet the peak demand; and no significant transmission constraints.

If these preconditions are fulfilled, market theory tells us that generators should be prepared to generate while the price is just above the cost of fuel, thus ensuring the lowest possible cost of power.

In this ideal market each generator offers a price at which he will generate. The system operator stacks the offers in price order and selects those needed to meet the load.

They are all paid the price offered by the highest-priced generator. This encourages everyone to bid in at his marginal cost and provides the extra income needed to earn a return on investment.

Clearly, the system works for base-load generators (such as the Otahuhu and Taranaki combined cycle stations) and the mid-range generators (such as New Plymouth and Huntly) which will be getting more than their marginal cost most of the time.

But how can the generators providing peak power and reserve capacity turn a profit?

Consider three generators each with 100MW units bidding for the last 100MW of load. They are competing to be selected to run, so they will all bid in at their marginal cost. Result: the one selected gets enough to cover his fuel cost but does not get enough to provide a return on his investment - and the other two get nothing at all.

So for generators providing the reserve capacity needed to make the market work, it is a loss-making business.

If there is no reserve capacity, competition disappears and generators are no longer motivated to bid in at their marginal cost. So the market fails, and in such a manner that the net result may well cost consumers far more than it was thought to save.

This is what has happened in the US and, to some degree, in Australia. The high prices are passed on to all generators; the base-load generators make huge profits while the peaking stations - the ones we need - run for only a few hours and get a tiny share.

So the market carries the seeds of its own destruction because it does not adequately reward those who provide reserve plant, and when there is a shortage, the market fails, prices skyrocket and it is no longer politically acceptable.

So how do these predictions line up with experience in the New Zealand market?

First, that base-load power stations are profitable. Three generators are competing to build a new 400MW base-load station.

Secondly, that holding stations in reserve for peaks or shortages is a dead loss. Marsden A and B are up for sale, gas turbines at Stratford and Whirinaki have been sold overseas.

Thirdly, that if there is a shortage, competition no longer controls prices. Prices go much, much higher than normal.

Thus there is a perfect fit between theory and practice.

Another question arises: does New Zealand, with only five significant generators and 70 per cent hydropower - which is subject to periodic droughts - fit the "market model".

It is easy to see that an all-hydro system does not fit. With hydro the cost of fuel (the water) is virtually zero but the capital cost is high. Also, many stations are "run of river" and will spill if they are not able to generate.

So "run of river" stations must bid in low and the stations with storage bid in at what they reckon the market will stand. Which most of the time is not much but during a drought is exceedingly high.

A business that loses money most years and needs to make huge profit during droughts won't attract investors.

Provided our mixed hydro-thermal system has adequate storage and more than sufficient thermal reserve capacity, the market will work.

But when we are short of water and all the thermal stations are flat out, the competitive price ceiling evaporates and it is impossible to predict how high the price will rise.

If the drought is severe, the hydro operators will bid higher and higher in a desperate attempt to conserve remaining storage. For this prudent action, they will be accused of manipulating the market. If they don't, and we run out of storage, there will be a national calamity. Either way, heads may roll - the wrong heads.

Whatever happens, the generators will be seen to be making windfall profits while their unfortunate customers pay high prices and are exhorted to reduce demand. But based on what has happened in Australia and the US, long before the demand reduces to match supply the price will be beyond a level acceptable to the public, the politicians and the economy.

And if that happens, what should the Government do? It should freeze the spot market, pay for all generation at about 5c a kWh and try to engender the unified national effort that got us successfully through the 1992 drought.

There is a need to go back to square one and find a better market model that makes sure that we have adequate capacity and recognises that, every 20 years or so, a drought may result in shortages that can be met only by a unified national effort.

This is an important and often forgotten consequence of us enjoying the low hydro power prices in between the droughts.

A return to centralised ownership should be regarded as a last resort. An obvious first move is to set up an independent market or mechanism for ensuring that we have adequate reserve capacity and no significant transmission constraints.

If we had an independent system operator, providing future reserve capacity could be his responsibility, just as there is now a market for hour-by-hour reserve capacity.

And transmission system constraints that also push up the price of power could be handled by having the independent system operator finance their removal if it reduces the overall cost of power.

* Brian Leyland is an engineering consultant.