By Mark Knight
June 3, 2002 — Losses are part of life in the electricity business. We use high voltage transmission lines to reduce losses, but we cannot eliminate them. Yet they represent a significant portion of generation. We can take account of losses in our calculations of system load by estimating them based on metering differences and sample measurements, but how accurate are these estimates, and who stands to lose or gain from errors in calculation?
In order to answer these questions, it is necessary to look at each of the areas in question, namely estimating losses in the transmission system, in sub-transmission, and in primary and secondary distribution; allocating losses to wholesale and retail energy trading, and by class of load; and, finally, determining the impacts of errors in estimation of these losses to determine who stands to gain or lose.
Transmission and distribution losses combined generally represent 6%-10% of generation. A retail supplier selling power in deregulated markets has to buy both generation and the associated losses. The problem is that there is not much room to build a competitive business in markets where the standard offer is low. The supplier cannot just increase its selling price to pass on the losses.
Instead, the supplier is faced with a tough market where the price dictates that costs need to be contained for the supplier to be competitive. It would be very appealing to the supplier if the supplier could improve its margins by paying for fewer losses. Also, if most of the suppliers retail customers have connections to the primary distribution system, or at a higher voltage, the supplier would expect the associated losses to be lower than the overall distribution loss factor, and that should be taken into account in loss estimation.
To account for losses that vary by connection or feeder requires the ability to accurately estimate load at a more granular level, perhaps even down to the customer level eventually. We fully expect over time that the estimation and measurement of losses will receive more and more attention as the participants that stand to gain from changes in method and accuracy lobby for such changes.
Especially noteworthy is the variation of losses with system load. Losses vary greatly with the system load (wire losses vary by the square of load, and transformer core losses do not vary once the transformer is energized). All too often, planners drastically under-calculate losses during peak times, when generation has its highest value. With transmission constraints and high demand causing volatility in wholesale markets this becomes a real issue.
For an average transmission and distribution loss of 8% of generation, losses at system peak can be approximately 12%, representing a large cost when generation can be priced at significant increases over normal rates during undersupply situations. In these situations, the marginal cost of supply can increase dramatically, to well above $1,000 per MWh. We are paying for a lot more losses in this situation, and we are paying a lot more for these losses.
At $1,000 per MWH, and 12% loss at peak, the effective delivered price of generation is $1,136 per MWH ($1,000 / [1 – 0.12]), nearly a 14% premium over the price at the generator bus bar. This would be the price paid by the retail supplier, whether or not the cost of the supplier is immediately passed on to the customer or spread out over time by an average retail pricing mechanism.
If the cost is not passed on to the customer, the retail supplier will eventually leave the market, at that time exposing the customer to the true costs. To this cost, the cost of transmission and distribution must also be added, and these costs may also be inflated due to the impact of losses. In fact, there is a double inflation at work. The transmission and distribution systems need to be sized up to handle losses at peak, raising capital costs per kWh, and the same ratchet effect as illustrated above for generation also applies, so that the delivered cost per kWh is higher yet.
One impact of deregulation and vertical disaggregation is the removal of incentives to minimize system losses. Time horizons for economic analyses have shortened drastically, and distribution and transmission companies are separated from the impact of distribution and transmission losses on generation costs. This is also true for other characteristics of distribution and transmission systems that lead to increased generation costs or reduced power output, such as power factor correction and power quality.
The supplier has a lot to gain from loss reduction and more accurate loss calculation and allocation, but may have little political leverage or financial clout to drive these changes. New incentive mechanisms need to be created to encourage economically appropriate conditions from the standpoint of the customer and general society. Perhaps this could be driven at the ISO / RTO level. If economic incentives or other appropriate voluntary mechanisms cannot be devised, regulatory fiats may be required.
Of course, the customer ultimately absorbs all costs of losses, and to the extent that losses that should be economically eliminated are tolerated instead, the customer is the loser.
Mark Knight has been involved in electricity market restructuring since 1989, helping companies to adapt to competitive, regulatory and data management pressures caused by deregulation. Formerly Knight worked for Cap Gemini Ernst & Young, Logica, and Andersen Consulting. Fred Plett has held senior positions in utilities and has worked for Logica, EPRI and AlliedSignal providing international business development, regulatory compliance and testimonials. For more information contact Knight at email@example.com or Plett at firstname.lastname@example.org
© Copyright 2002 Mark Knight and Frederick Plett.