RiskAdvisory of Calgary, Alberta is a consulting and software firm that helps mid-size utilities get a handle on the many uncertainties that underpin every aspect of their operations--how much energy to produce, when to produce it, and how to price it. The software development side of the company, which is soon to achieve status as an independent company, provides customers with a fully integrated risk management package that includes front, middle, and back office functions. In the now-deregulated environment in which the utilities compete, accounting for risk in all these operations is a tall order and requires mammoth simulations.
Parker Sjogren came to RiskAdvisory to assist the development team in creating the new Monte Carlo component of their Energy Bookrunner application -- software that can efficiently accommodate mountains of detailed data on power generation, deals, and market conditions using both sampling logic and business logic. The resulting "Earnings-at-Risk" information that the software provides is used by utility clients to guide many complex decisions.
Risk Management Specialist, RiskAdvisory
To effectively take into account all the required risk factors, RiskAdvisory turned to the @RISK Developer Kit (RDK) for its powerful Monte Carlo simulation engine. When Parker first began experimenting with the RDK to incorporate the simulation model into the Oracle based system, the software ground through the huge wall of data in five days running on a high-end PC. Though impressive, this did not meet client needs for up-to-the-minute information. So Parker consulted with Palisade technical staff and began using the new RDK 4.0. With strong technical assistance from Palisade, the five-day wall was reduced to run the huge models in under 3 hours - more than forty times faster than before!
Mike Busby, the risk management specialist who works directly with the utilities executives, says this new speed is essential to utilities decision makers. In less than three hours, the software can process historical data on generator portfolio and load requirement for a plant with 18 generators, running 10,000 iterations for each month of a given year, and project a distribution of outcomes. The ability to see literally tens of thousands of possible scenarios - and their likelihood of occurrence -- with so many diverse variables has proven invaluable. In the tumult of today's energy markets, Busby says, "It helps people sleep at night."