Life on the Margin: What Generators Need to Know

March 2011

• A clear understanding of true marginal economics will help owners and operators of gas-fired generation maximize the value of their expensive assets.
• The situation calls out for a well-considered dispatch model that incorporates accurate estimates of the key parameters to enable the operator to quickly and accurately assess the run/don't run decision. But such a model needs to be continuously updated for changes in gas prices, HOEP, Global Adjustment values, heat rates, gas and utility distribution rates, and other parameters.
• Seems like a lot of work, but the work is necessary. The only thing worse than not running the generator and losing out on a chance to make money, is running it when it should not be running and incurring an operating loss.

Ontario's natural gas and electricity markets are both complex and volatile markets. This poses a challenge for operators or natural gas-fired power generation facilities. Whether you are a district energy operator or your institution or industrial operation has embedded generation, your operating economics depend directly on understanding what is happening in those two markets, relative to one another, hour by hour. That is a daunting task!

Usually, the owner-operator of generation has the freedom to decide at any point in time whether to run the generator or not. The driving force for this decision would normally be the economics - run if it is profitable to do so. This comes down to a comparison of the marginal cost of running versus the marginal revenue earned from doing so. How much will my costs increase if I run for one more hour? How much money will I make (or costs will I avoid) if I run one more hour? But these questions are more easily asked than answered.

Looking first at the revenue side, and focusing on generators that are not dispatchable by the Independent Electricity System Operator (IESO), the starting point is the Hourly Ontario Energy Price (HOEP). A generator selling power onto the grid would typically earn HOEP on each MW of output. An embedded generator (where the generator output is less than the gross load of the operation) would displace purchased power at the HOEP price. HOEP is calculated by the IESO as the average of the twelve 5-minute Market Clearing Prices for the hour. Importantly, HOEP is known only after the hour is over. But the decision to run is made before the hour starts. So the operator must anticipate what the HOEP will be and project revenues in order to make the run/don't run decision. Sometimes, pre-dispatch prices are a good indication of what HOEP will result...but not always. As they say in advertising, "actual results may vary"!

For embedded generators, it is important to also take into account other costs that may be avoided on the margin. For example, generation that reduces the consumer's net power load may produce savings on items such as transmission and distribution, Global Adjustment, wholesale market service and Debt Retirement (grandfathered projects) charges, and local line losses. Running embedded generation in just a few peak hours can reduce demand charges for distribution and transmission services for the whole month, and could also affect the proportion of Global Adjustment charges allocated to that consumer over the whole year. The impact of these savings could outweigh the influence of HOEP on the operating economics.

On the marginal cost side, the biggest expense will be natural gas costs. Understanding the marginal cost of the gas required to run one more hour is complicated by a number of factors related to utility rates and services, and how the operator's natural gas purchasing arrangements are structured.

Some utilities offer both "bundled" and "unbundled" distribution services. The key difference for the power generator is that unbundled services involve closely matching the volume of gas delivered today to the volume of gas consumed today. Bundled services create a bigger disconnection between the timing of the buying and the consumption.

A closer connection between buying and consuming is both a burden and a blessing to the generator. It is a burden because it requires more work to manage gas supply each day to match supply to consumption. It is a blessing because it enables a direct implementation of the run/don't run decision. Under an unbundled supply arrangement, the cost of a small daily increment of gas (enough to run today) will be clearer to determine. And if I can't run profitably at today's power prices and gas prices, then I don't buy the gas.

In contrast, under a bundled distribution service, I will buy gas today whether I run or not. I need the utility's permission to stop the flow of gas supply in my contract. If I must buy gas today but I don't intend to burn it, then I must hope it will be economic to burn it at some later date (or, in the alternative, economic to resell the gas at a later date if I never burn it). Or, if I run the generator today and use a lot of gas, some of that gas may be bought at a later date, hopefully at a price low enough that it was profitable to burn it.

In addition, the distribution rates charged for unbundled services typically involve mostly a monthly fixed cost, and a very small volume-related charge or none at all. The marginal cost of incremental distribution service is clear. Bundled distribution services often involve billing on block rates, where the cost of the last unit of gas distribution services in the month depends to some degree on the total volume of services used in the month. In order to assess the cost of using one more hour's worth of gas, I must first anticipate how many hours worth of gas I am likely to use this month.

The operator must also be aware of the other marginal costs of operation. In some cases, engine or turbine maintenance contracts assume work to be done at various increments of running hours. It makes sense to amortize these costs on a per-operating hour basis.

In addition, the operator needs to accurately understand the facility's heat rate under the applicable seasonal operating conditions. An accurate estimation of heat rate depends on understanding alternative boiler efficiencies, an area where imprecise assumptions can produce faulty conclusions about operating economics.

The situation calls out for a well-considered dispatch model that incorporates accurate estimates of the key parameters to enable the operator to quickly and accurately assess the run/don't run decision. But such a model needs to be continuously updated for changes in gas prices, HOEP, Global Adjustment values, heat rates, gas and utility distribution rates, and other parameters.

Seems like a lot of work! But the work is necessary. The only thing worse than not running the generator and losing out on a chance to make money, is running it when it should not be running and incurring an operating loss.

A clear understanding of true marginal economics will help owners and operators of gas-fired generation maximize the value of their expensive assets.

Combined Heat and Power Projects: An Opportunity But Know the Risks Read more »

Combined Heat and Power Standard Offer Program: Managing Gas Cost Risks - Presentation by John Voss, Managing Director, Aegent Energy Advisors Read more »