Volatile energy markets are pushing some facilities to use wholesale market pricing forecasts to help develop retail energy budgets. Such forecasts are found as futures pricing (e.g., NYMEX), forward market projections (e.g., Platt's), private energy consulting firms (e.g., Enercast), and government sources (e.g., DOE, ISOs, PUCs). While not always accurate, they provide quantitative information useful for annual energy budgeting.

Retail energy prices may, however, be only indirectly related to wholesale pricing, or may lag behind the wholesale market, making short-term projections speculative. The usefulness of such predictions depends on one's sensitivity to wholesale market pricing and his expectations for forecasts of them.

Central Factors

While various issues enter into the discussion, the most important are:

  • How a facility buys its energy (e.g., through a utility or marketer);
  • Where it's located relative to the nearest wholesale purchasing point (i.e., a "hub"); and
  • The facility's energy usage patterns (e.g., how its consumption varies with time).
Wholesale pricing and forecasting data become more useful as a facility's purchasing methods and usage patterns approach those of a small utility or wholesale energy trader (e.g., sensitive to market fluctuations but having a high load factor).

Market Distortion Mechanisms

Only major energy users (e.g., industrials, very large campuses, and utilities) buy directly from wholesale suppliers. The rest of us buy from utilities or retail marketers. When buying from a utility, wholesale pricing is just one part of the retail price developed under a regulated tariff. If buying from a retail marketer, a private contract may instead control energy pricing (though delivery cost may be subject to a tariff).

Such agreements mute or distort the impact of wholesale pricing. Utilities may, for example, absorb large short-term price swings and then recoup them over a much longer time period, perhaps through a later general rate increase. Some retail energy contracts and utility tariffs contain mechanisms (e.g., daily/monthly balancing, fuel adjustment clauses) that pass price fluctuations through to monthly utility bills. The value of a wholesale price forecast may depend on the type and degree of such pass-throughs.

Depending on the location at which a price is forecasted, the cost of interstate transportation (i.e., "basis") may magnify that impact. That charge varies greatly with seasonal demand and pipe/power line availability, and must be added to prices forecasted for major pricing hubs (e.g., Henry Hub in Louisiana). Any forecast that does not include basis may have only marginal value.

Unlike many process-based loads (e.g., manufacturing) that are unaffected by weather, most facilities use energy based on occupancy and temperature. Their load profiles rarely look like the neat blocks of energy priced in wholesale markets. Approximating a retail load profile using smaller blocks of various size and duration is possible, but may significantly raise the final price.

Energy traders and utilities use financial and other mechanisms to mute the impact of annual variations on their bottom lines. To some, each month is like a poker hand: a player may occasionally lose his shirt but, if he plays long enough (i.e., uses probability-based price forecasts), he's bound to break even. Few customers have such latitude, limiting applicability of short-term forecasts. If, however, one is looking at a multiyear contract or for only general guidance on price direction, such forecasts may be useful.

Making Forecasts Relevant

Before trying to develop price forecasts, find out if your utility, ISO, marketer, or PUC have already done this task (perhaps as part of a recent rate filing). Be careful with such data: it may be dated by the time you read it, or may represent the total impact on a utility rather than the pricing to be seen in a fuel adjustment charge.

Review your tariff to find that part of a facility's energy bill directly affected by wholesale pricing (e.g., the monthly fuel adjustment charge). Historical data for that monthly charge may show its location and seasonal sensitivity. Use data from recent years when weather was relatively normal (i.e., close to the 30-year climate average).

Using past wholesale pricing for the same months at the nearest pricing point, calculate the differential impact due solely to wholesale variations (i.e., no tariff changes during that period). Add those differentials to the prices found in a wholesale monthly forecast (for that same pricing point) to see how the fuel adjustment charge may vary in the coming year. Add those numbers to the other components in the tariff pricing (e.g., delivery, distribution, etc.) to create a complete price curve. ES