Why WEMDG made the wrong choice

The wholesale market was launched over a decade ago to deliver electricity at the lowest cost to the economy. Bryan Leyland, from Leyland Consultants New Zealand, compares it with the ‘single buyer’ model alternative and finds it lacking.

The Wholesale Electricity Market Development Group (WEMDG) designed the NZ electricity ‘competitive’ market, launched in 1997, to “ensure that wholesale electricity is delivered at the lowest cost to the economy”.

The alternative ‘single buyer’ (SB) model was rejected by WEMDG. Under this model, the single buyer is independent of the both the government and the industry, and an independent auditor handles any complaints. The single buyer manages the system operation using the traditional and proven ‘stack’ of generation based on increasing marginal cost and transmission losses. They would also monitor hydro lake levels, and schedule lowest-cost thermal generation to run flat out during the summer if needed to build up the storage. The SB then sells wholesale electricity using cost reflective tariffs. As electricity generation is a long-term business, a single buyer must make a range of estimates of likely future load growth and then evaluate all technically feasible and economic resources available for generating electricity. Each load growth scenario will have a total cost of generation associated with it. A range of generation options will be needed because, until tenders are received, there will still be major uncertainties in the costs of the various projects.

Based on the preferred options for generation development, the single buyer can ask for firm offers for the design, construction and operation of a range of power stations.  Calling tenders of new generating capacity taps into what I believe is the only truly competitive market in the electricity industry – competing for long term contracts to build, own and operate power stations. The single buyer option is, in my view, most definitely a market-driven option in a market where there is price elasticity and many alternatives.

The various offers would then be compared one with the other on the basis of long-term cost, security of supply and the associated transmission costs and losses. Contracts would be placed with the selected tenders. The contract payments would consist of a fixed payment representing a reasonable return on the capital cost of station, and variable payments for the amount of power actually generated. Structuring payments in this way means that the system operator can schedule more thermal generation in a dry year and less in a wet year, without seriously affecting the cash flow of the owner of the station.

To estimate the cost of electricity if a SB option had been adopted, I have assumed that the ECNZ power stations were sold off at their book value – $800/kW. This is based on the $4.5 billion valuation when Transpower was split off from the ECNZ¹.

A spreadsheet has been used to calculate how much the generation from these ex ECNZ stations would have cost between 1997 and 2007. Reasonable estimates have been made regarding the cost of gas for gas-fired power stations and for enhancements to the hydro power schemes. According to my calculations, the generation from the ex ECNZ stations would have cost a single buyer $14.5 billion over the 11 years from 1997 to 2007.

My spreadsheet also calculates the cost of generation from the schemes built since 1997 (‘new stations’). It is based on reasonable estimates of operation, maintenance and fuel and capital costs and indicates that generation from new thermal stations would have cost a single buyer about $2 billion, from geothermal stations about $300 million, and from wind power stations about $210 million.

With a SB there would also be extra costs involved in running the operation. A reasonable estimate is that this would cost $10 million more per year than the present cost of running the Electricity Commission and associated organisations such as M-Co.

When all these costs are totalled, the cost of electricity using a single buyer market would have been $17 billion.

I have also calculated the actual cost of wholesale electricity over the same period. The costs are derived from the average spot prices published by M-Co, multiplied by the annual generation used for the single buyer model. This calculation carries an assumption that, over the period, the cost of electricity on the spot market was much the same as the cost of electricity sold on the contract market. As hedging is widely believed to be a ‘zero sum game’, this is a reasonable assumption².

If the system had been centrally coordinated and optimised, we can be reasonably confident that the new 400MW combined cycle station at Huntly would have been built in Auckland instead of 90 kilometres to the south. Savings in transmission, losses etc could have been $350 million over the eleven-year period.

With a single buyer regime the electricity shortages of 2001, 2003 and 2006 and 2008 would have been less severe and the spot prices would have been under control. This would have reduced the economic loss from industry and commerce being forced to limit production. I have assumed that at least $200 million of economic loss can be attributed to the ‘market’.  

The cost

Taking all the above into account, the calculation gives a total cost of $23.5 billion. This is $6.5 billion more than the single buyer model. I believe that this is a conservative estimate of the extra costs that the electricity market has imposed on New Zealand’s electricity consumers, and I have not tried to calculate the cost of transmission constraints, which I am sure is huge.

We have paid between 25 percent and 50 percent more for wholesale electricity than they would have done if the Wholesale Market Development Group had chosen the single buyer option.  

When the electricity market was first proposed, I asked one of its proponents how it would cope with dry years. Their reply was “in dry years the price will go up and demand will go down: we will not have a problem.”  It seemed to me that this response was based on blind faith in the market, ignorance of the fact that there is very little price elasticity in the electricity market, and a very poor understanding of the realities of electricity supply in New Zealand. 

All electricity markets around the world that are similarly structured to the New Zealand market are ­– or soon will be – suffering from a lack of capacity. According to Boston Consulting Group³ it is inherent in the electricity market model that there will be regular shortages - an endless series of “boom and bust”.

So how and why did they all get it so wrong?  Firstly, I suspect that no one in WEMDG asked the critical question: “Should electricity be sold at the average cost of generation or should it be sold at a cost related to the marginal cost of generation?” 

Economists argue that it should always be the marginal price. When dealing with most market commodities, the most modern factory usually has the lowest cost of production so marginal cost pricing serves to drive out old, inefficient factories. As it should. But in New Zealand, the lowest cost production is from our hydropower stations. The actual cost of production is quite low - probably less than 2c/kWh. So the direct – and predictable – consequence of introducing the WEMDG market model is that old hydro stations will return huge windfall profits. 

We all accept that we get roads, water, sewage at the averaged price, or something close to it. Why not electricity? 

In conclusion I believe that WEMDG’s belief that “the market will deliver” should have been tempered with wisdom and commonsense. Markets work when the commodity being traded has price elasticity and an alternative good. Electricity has neither.  Economics 101 could have told WEMDG that the market model they selected was fatally flawed and that has been proved by the fact that, by a large margin, wholesale electricity has not been “delivered at the lowest cost to the economy”.

1. At that time, many people believed that Huntly Power Station was virtually worthless and would soon be shut down.
2. In reality it is more likely that it under-estimates the cost than it over-estimates it.  Consumers who choose to buy an electricity hedge are choosing  to buy insurance and this carries its own costs.
3. “Keeping the Lights on: Navigating Choices in European Generation” Boston Consulting Group May 2003.