ESB customer supply implements LODESTAR for demand forecasting

ESB Customer Supply is the largest supplier of power to final domestic and business customers in the Irish market, with 1.9 million electricity customers (over 97% of the market) with annual consumption of about 14 TWh (60% market share). With the reviewed design of the Irish electricity market and the All-island Single Electricity Market Project planned for July 2007, ESB has been preparing for the changes in its external wholesale environment.

The company identified demand forecasting as a key IT component to structuring their IT systems, integrating customer portfolio management and energy sourcing management. This would serve as a cornerstone to enable them to evolve with the new market and to support their activities within the future electricity pool.

SYSTEM REQUIREMENTS

 ESB Customer Supply needed a short and long-term vision of their customer portfolio, their associated consumption and energy purchase forecasts, and consequent revenue streams. Therefore the company required a forecasting system that would generate and maintain short and longterm forecasted customer consumptions and customer numbers by tariff. The system also needed to allow for forecasts to be rolled up to the tariff group, market segment, for the entire ESB Customer Supply organisation, or for the entire Republic of Ireland.

ESB required the consumption forecast to be driven by interfacing to external parties and to historic data maintained in other ESB Customer Supply systems. The historic data needed to be applied to a number of long and short-term forecasting models, with the resulting forecasts being accessible through a comprehensive suite of reports. An imperative requirement was that the systems would enable ESB Customer Supply to generate forecasts for both the whole Republic of Ireland market and their own market share and compare them to the contracted power purchases.

The forecast models which were adopted at ESB comprised regression, trend, like day and proxy day using consumption, customer numbers, temperature and GDP as inputs. The company also needed customised routines for the system to adjust its forecasts based on system operator published load forecasts, contracted customers additions and losses, or simply based on manual inputs.

ESB Customer Supply found that LODESTAR met and exceeded all of these requirements. “LODESTAR’s proven flexibility and robust forecasting functionality enables ESB to be more successful in the current market and positions us for operation under the future pool market,” asserts Ciaran Doran, Programme Manager, Wholesale Capabilities Development, ESB. “After evaluating a large number of other vendors, we found the LODESTAR solution to be the most suitable to address our complex needs.”

FAST IMPLEMENTATION

LODESTAR® Customer Choice Suite™ (CCS™) was implemented in six months. A major component of the implementation was the configuration of LODESTAR® Adapter™. This was done to handle interfaces to external ESB Customer Supply data sources including historic and forecast data, scalar and half-hourly metered consumption, customer numbers and load profiles, actual and forecast temperature received daily from Met Éireann, market settlement consumptions, and energy purchase contract data.

A key aspect throughout the implementation process was the excellent relationship that developed between LODESTAR and ESB business and project teams. This relationship guaranteed that the implementation not only delivered business functionality, but also focused heavily on operational use and analysis of the forecast modelling output.

LODESTAR CCS was implemented on time and within budget, enabling ESB Customer Supply to generate their first automated demand forecasts starting in January 2006. The forecast provided the ESB Customer Supply forecasting team the opportunity to not only review the high-level results but also the underlying tariff level data used to generate the forecasts. As a result of their initial reviews, the team was able to take advantage of the inherent ease of use of the system. In particular, the LODESTAR Adapter interface enabled them to easily update the historic and standing data, such as demand profiles, and re-initiate long and short-term forecasts.

LODESTAR CCS is a flexible forecasting solution to generate strategic views of the customer portfolio and the corresponding energy requirements, integrating strategic information on customers, markets, and electricity consumptions available in ESB Customer Supply. The solution is also very flexible and can be easily configured to use further data as it becomes available. An example was the enhancement made shortly after implementation at ESB, where wind generation data was incorporated to provide a more accurate measure of the system demand. The single electricity market due in 2007 will undoubtedly provide access to a greater range of market data which could be easily incorporated into the system.