Two states have decided to review the high cost of gas this past winter and the effect the price hike has had on the states' gas utilities.
Florida. While granting requested rate...
the new load shape, which could prove more expensive to serve.
The three strategies explored below remove customers from a rate class load profile using as selection criteria
(1) size, (2) load factor and (3) size weighted coincidently with class peak. The removals are simulated for residential, small commercial and larger commercial rate class profiles. The point of this exercise is to see how retail vendors could exploit opportunities induced by the practice of load profiling as markets develop and progress.
The valuable trick will be to employ commercially available databases to develop and execute real world marketing strategies. The database will identify customers with better-than-average load characteristics and remove them from the load profile class via more precise metering.
But how can retail providers identify the best customers? To complicate the problem, power pools, control areas and independent system operators will likely develop a diversity of cost-causality rules. In New England, for example, it appears that capacity costs will be based on a retail provider's individual monthly peak, independent of the regional peak. If this cost-causality rule is employed, the good customers will be those that tend to be less coincident with whatever loads the retailer is already serving. If another rule applies, the definition changes.
To deal with this diversity, any successful retail provider targeting smaller customers must be familiar with (1) the load research and delivery tariffs of the distribution utilities within whose service territory it wishes to operate, (2) the cost causality rules of the relevant control area and (3) the service terms mandated by state regulatory commissions.
Once the retailer has a firm grasp on these considerations, it will stand ready to try to arbitrage the class-estimated load shapes with a particular customer's actual metered load: Let the cherry-picking begin.
Three Strategies for Exploitation
At Central Vermont Public Service Corp., we explored what might happen if a load-serving entity systematically identifies, procures and serves customers by removing them from the load profile by installing interval metering. In particular, we tested three strategies: %n5%n
Strategy 1. Choose customers with the highest individual load factors.
Strategy 2. Choose customers with the highest demands. (These customers likely consume the most energy and have the most potential for sales growth if their load factor improves.)
Strategy 3. Choose customers that add a relatively small capacity requirement contribution per kilowatt-hour of sales to the LSE's total capacity requirement.
(Since, for the purpose of this experiment, the LSE's other loads are unknown, I assume that the LSE's total load shape looks just like the load shape of the profiled class. Since the rate classes analyzed here are uncontrolled, and not priced at time-of-day rates, a well-diversified LSE certainly could have a load shape similar to the load shapes of the classes profiled in testing these three strategies.)
These three strategies are applied to four profiled rate classes: GD (em commercial, general service demand and energy rate; G (em commercial, general service without a separate demand charge; D (em residential, applicable to the vast majority of residential customers; DH (em residential, "total electric living"