TO:       Members of Load Profiling Working Group
FROM:     M.L. Chan, ML Consulting Group
DATE:     September 22, 1998
RE:       Comments on The Posted Questions for 9/25/98 Workshop
________________________________________________________


With respect to the questions to be discussed in the referenced workshop, I have the following
comments:

Question #1 - What are the costs associated with hourly interval metering?
Comments -     The cost would involve a headend computer master, a telecommunications link (which
               may be composed of a WAN and many LANs), and many end-use solid state meters. 
               The functionalities of this complete metering system depends on the portfolio of
               energy and value-added customer services being offered to customers.  Just metering
               alone would make the system cost prohibitive.  However, combining all those
               functions could render the system cost effective, so that metering does not have to
               shoulder the entire cost of the system.  However, the cost of the communications
               system may also have to be higher because of the larger bandwidth requirements if
               some data-intensive value-added services are to be offered.

Question #2 -  What is the likely cost impact of imposing the hourly interval metering requirement
               on customers whose maximum demands fall within the 20 to 50 kW range?
Comments -     The cost impact may depend on who owns the data and the meters.  If the customers
               own the data, they can sell the information to marketers if they so choose.  Then the
               resulting revenue could help pay for the meters.  But there is also the question of how
               to allocate the cost of the common equipment (i.e., headend computer and the
               communications system) to different customers, which will determine the cost impact
               to them.  Besides, customers may need more than just the load data, but also the load
               shape forecasts if they intend to optimize their decision on which energy service
               providers from whom they want to purchase their energy.  Therefore, the "metering
               devices" may also have to have load forecasting and transaction evaluation
               capabilities, though such capabilities could be available on an Internet-based service
               provider.  This would imply that such customers would need to have access to the
               Internet.

Question #4-   Should the load profiles for these customers be extended or discontinued in light of
               the metering situation?
Comments -     I believe that load profiling offers an interim and cost effective solution for the
               problem at hand.  It also provides a good starting point for developing the eventual
               load shape data base that will need to be dynamically updated.  The metering system
               could be easily extended to provide such a dynamic update of the data base.

Question #5 -  Should hearings be held to resolve this issue?
Comments -     I believe that the Working Group should sponsor a study to define the statistical
               nature of the inconsistencies and offer solutions to resolve the problem.  The
               inconsistencies may have to be compared against the data accuracy requirements as
               specified by the load profiling approach to decide whether refinements are justified or
               needed.