Flick switch up. Light goes on. Flick switch down. Light goes off. Flick switch up. Light goes on….and so it continues. This is how my four year old influences the energy market, completely unaware of the growing list of companies competing to charge me for the energy she is so eagerly consuming.
Since the 1980s, Governments throughout the world have embarked on the process of deregulating all- or part- of their energy networks in an effort to improve the quality and service of energy delivered to consumers, and provide transparency of power pricing. I’ve worked with customers that operate throughout the electricity supply chain – Generators, Distributors & Retailers – but none interest me more so than those termed “Gen-tailers”, that is, companies that both generate electricity and sell it to the consumer.
The energy market is unique – the item being traded cannot be kept in stock, customers do not pre-order, and both demand and supply can vary significantly. As a result, market prices are highly volatile at times of peak demand and supply shortages, and expose companies to significant financial risk.
Operating both upstream and downstream of the energy supply chain, Gen-tailers are in a unique position to make profit both when prices and low and when prices are high. Put simply, in times of low prices Gen-tailers can choose to generate less electricity in their own stations, purchase additional units on the spot market, and make profit selling to their customers. In times of high prices, Gen-tailers can choose to generate more, sell additional units on the spot market, and make profit upstream.
Which begs the question: How much should we generate?
Gen-tailers, like all generators operating in a deregulated electricity market, need to commit in advance how many units they are planning to generate. This enables the regulators to ensure there is enough energy supply to meet expected demand. Depending on the market, this commitment is often required days in advance. Traditionally, this unit commitment has been set based on gut feel and historical results. Given it’s critical impact on profitability, more and more companies are looking to analytics to evaluate different positions and support this critical business decision.
To make an informed decision about unit commitment, here are the key questions we need to answer for any given period:
- How many units will my customers demand?
- What will the spot price be?
- How many units can I generate?
Using IBM Business Analytics, we can predict both expected customer demand and spot price, taking into consideration historical behaviours, weather forecasts, event calendars and network conditions for example.
These predictions can form the baseline in our “what-if” scenario modelling application where we can make manual adjustments based on events and information we know are not already captured in the data. We can evaluate different scenarios within the application to understand the financial position for a specific time period – for example, what if we commit to generating 80% of our customers’ demand and buy the shortfall on the spot market? What if we commit to generating 120% of our customers’ demand and sell the excess on the spot market?
Our financial position is also impacted by our generation cost – we can use IBM Business Analytics to optimise the generation of our units. Taking into consideration planned downtime of facilities, start up costs, hourly running costs and generation capacity, we can determine the optimal spread of generation across all available facilities for that period, and incorporate the projected operational cost into our Unit Commitment scenario to fully understand our financial position.
What you see on the glass, is a simple, easy-to-use analytical workspace where an analyst can evaluate a range of commitment scenarios with understanding of what is likely to happen in the future.
What you can’t see behind the glass, is a suite of advanced technologies, pre-integrated to deliver analytical insight to the business – specifically IBM Cognos Business Intelligence, IBM Cognos TM1, IBM Cognos SPSS and IBM iLog CPLEX.
Flick switch up. Profit goes up. Flick switch down. Profit stays high. Flick switch up. Profit goes up…..