Modeling energy transactions in deregulated electricity markets

Similarities between the Californian and European power markets suggest multidimensional modeling approaches for market participants in Europe are appropriate because they face the same market conditions. Henry Shin and Dr Lawrence R. Conn explain.

Henry Shin & Dr Lawrence R. Conn, USA

Today’s energy markets are changing fast. This fact is not news, but it does raise the question of how firms are to deal with the rapidity of change. European power markets are in a state of flux from a regulatory perspective, with new deregulation initiatives and policy imperatives such as carbon emission caps. How does a company model the impacts of these multidimensional changes in a timely manner? Given the rapid speed that these new demands are imposed and the complexity of interrelationships between new market requirements, modeling efforts must keep pace with these increasingly complex and quickly changing markets.

Click here to enlarge image

A modeling approach that offers flexibility, transparency and dimensionality in implementing highly complex business rules is necessary to accurately represent the business environment and support sound decision-making. Multidimensional business modeling technology can help companies integrate business processes to account for new demands on energy markets à‚— something that is especially important for the small-to-mid-sized firms that cannot afford enterprise-wide systems and rely mostly on spreadsheets to meet their analytical needs.

Supply Chain Puzzle

Because electricity cannot be stored, it must be generated upon demand, delivered to the right buyer at the right time, and meet transmission constraints. This creates a gigantic supply chain puzzle. All of these transactions must be scheduled with the jurisdictional power grid operators such as Elexon in the UK or NEMMCO in Australia. In California, the California Independent System Operator (CAISO) operates the spot markets for power and settles the transactions.

After delivery of the power, the trading office needs to estimate all the charges and payments to evaluate its financial positions on a daily basis. Once actual generation and usage data are available, the financial settlement process requires accurate and timely calculations by the back office à‚— essential because market participants have a limited time to reconcile their transactions and dispute their charges.

Add constantly changing business rules, and the task becomes even more daunting. Power market participants working in evolving markets must be prepared for frequent changes to their financial models.

Market redesign in California

What does market redesign in California have to do with the international electricity industry? The EU published a package of energy proposals last September that bear remarkable resemblance to the proposed market structure in California. Among other proposals, two of the most significant are unbundling the operation of generation and transmission (i.e. the use of an independent system operator model for grid operations and open access to the transmission system) and greater transparency into energy transactions.

California is alone in deregulating its energy markets in the Western US. This creates ‘seams’ issues with control areas in surrounding states. This situation is directly analogous to the situation in the European Union where member states are in various stages of deregulation. Additionally, as with many European countries, California is in the process of adopting carbon emission limits. Because the design of the carbon markets is unfinished, a lot of uncertainty exists among market participants.

A notable feature of California’s new market design is the use of locational marginal prices (LMPs) at approximately 3000 points (or nodes) to balance the system supply and demand at the least total system cost while meeting operational and security constraints. Based on this nodal design, CAISO will add a forward market (Day-Ahead Market) and transmission rights (Congestion Revenue Rights) markets, and redesign the Real-Time and Ancillary Services markets that it currently operates. Significant changes to the technology have been made to support the additional complexities and expanded financial products. ‘Bid-to-bill’ processes to complete an energy transaction in the California market have become more complex in this new environment. These changes add up to a whole new way of doing business for market participants with respect to trading, scheduling and settling energy transactions, creating special challenges for small-to-mid-sized market participants that may not have the resources for an enterprise-wide energy transaction infrastructure.

Deregulated energy transactions

A typical ‘bid-to-bill’ process in the deregulated energy market involves submitting bids to sell or buy spot energy and ancillary service products, scheduling all market and contract transactions, and finally settling financially with the market operator. The front office often maintains daily estimates of its financial positions and provides monthly estimates to the accounting group for month-end closings. The back office has the responsibility of validating the settlements from the market operator and disputing any discrepancies. It is also often responsible for settling internally between different business units or generation facilities, requiring further allocations and assignments of charges and payments. Once the settlement statements are available from the market operator, the estimates the front office sent to the accounting group are replaced with the actual settlements from the back office.

Market participants need accurate settlement information to maximize the value of their energy portfolio and to manage transaction risks. Settlement data can be used to develop bidding strategies and back-test them after energy has been delivered. They are also essential to minimize avoidable financial penalties levied by the market operator. As the California market has shown, newly emerging deregulated markets will likely experience a number of market data changes and settlement reruns. A robust settlement validation and disputes capability certainly offers a compelling value proposition for any participant in these deregulated markets.

Despite the value that accurate settlement capabilities can add to an organization, integrating them in the entire chain of the ‘bid-to-bill’ process is difficult for many organizations. Some market participants have implemented a class of enterprise-wide tools called energy transaction and risk management (ETRM) systems, but small to mid-sized market participants frequently rely mostly on spreadsheets to meet their settlement needs.

Investment in information Technology infrastructure

Financial settlement of deregulated market transactions typically involves allocations and assignments of market costs and revenues using billing determinants that are based on cost-causation principles. Calculating all the billing determinants that go into final billable quantities for each charge is often very complex, involving numerous variables with different combinations of dimensions such as time, resource type, transaction type, location, and other attributes. Due to the complex data relationships, two-dimensional spreadsheet-based templates are inadequate to meet the demands of deregulated settlement processes. Plus, auditing spreadsheet calculations to ensure conformance with the market operator’s specifications and updating formulae as business rules change are formidable challenges for any market participant.

Transitioning to the new market processes presents another challenge: adapting existing technologies to handle the new technological standards. For example, a file format based on eXtensible Markup Language (XML) is coming into common usage. The XML-based transaction files represent a great improvement in terms of computers exchanging information efficiently, but the new files are virtually impossible to decipher for business analysts because of complex data relationships and volume of data. Without significant investments in information technology (IT) infrastructure and capabilities, adapting existing technologies may not be a viable option.

Managing complex settlement demands

An example of new technology being introduced coincident with new markets will provide a good example of the complexity involved. In California, the CAISO will be providing an enormous volume of data for market participants. The new market design will have at minimum three issuances of settlement statements for each trade date, along with a bill determinant file. The latest test version of just one bill determinant file exceeded 80 megabytes, with 1.2 million records. Spreadsheets will not be able to manage such a large and complex volume of data.

Databases can easily accommodate these files, but the development and maintenance of databases to calculate model scenarios may exceed the IT resources of many small and mid-size market participants. Today, new modeling software offers the technology that companies need to manage these complex settlement demands. Power Market Consulting Incorporated, for example, has developed an analytical tool designed to validate CAISO’s new charge codes based on exact formulas specified in CAISO’s Settlements Business Practice Manual (BPM). The settlement solution, Rozetta, is based on multidimensional modeling technology developed by Quantrix, and is available either as a standalone desktop application or is used as a front-end to an XML database that contains CAISO transaction files in their native format.

Multidimensional modeling software directly addresses the shortcomings that most common settlements methods and tools currently have with new approaches to energy transactions and market clearing. Modeling software:

  • Provides the ease of use of spreadsheet technology with the power of database technology
  • Provides the optimal balance for a dynamic market with a lot of data
  • Is very robust and can handle virtually any amount of data
  • Is limited only by the operating system and hardware constraints of the user’s equipment

Spreadsheet technology relies on physical cell locations in an x-y grid. It is extremely difficult to alter the data structure in a complex model without breaking the dependencies, a nightmare in terms of implementing any changes in the model. With as many formulae as the number of calculated cells, it is very hard to get a clear picture of what each cell represents, leading to potential audit issues. This approach is particularly limiting when determining energy settlements, as companies need to review multiple data points from a variety of perspectives, quickly and easily.

Multidimensional modeling technology enables companies to represent all the variables à‚— inputs, intermediate billing determinants and ultimately the settlement outputs à‚— in matrices by their attributes. The data can be presented in a different layout by simply moving the tiles (attribute labels) around the matrix. Spreadsheets simply do not have this kind of capability or the flexibility to capture and work with multi-dimensional data. Complex settlements formulas that define the mathematical relationship between input and output variables are modeled with ‘write-once’ formulas. Unlike formulae in spreadsheets, the modeling software’s formulae are not tied to physical cells and need not be repeated as copied formulae. The robust approach to computations makes complex settlements formulae much easier to develop, audit, and maintain.

Proprietary database applications can perform complex calculations, but constant programming and maintenance means they are costly and often aren’t nimble enough for changing market rules. Spreadsheets, on the other hand, cannot handle the vast quantities of data and multiple dimensions, and become unmanageable, prone to formula error, and unworkable when incorporating new business rules and functions. A settlement solution using multidimensional modeling technology fills the gap between databases and spreadsheets, especially for small and mid-sized organizations. Low cost, ease of use, power, and flexibility make multidimensional modeling software the tool of choice.

Henry Shin is the founder of Power Market Consulting, an authorized partner of business modeling and analytics software provider Quantrix, and has been providing regulatory and business systems consulting services since the start-up of California’s deregulated power market. He has extensive experience in the development and implementation of energy settlement systems with transaction values in multi-billion dollars.

Dr Lawerence R. Conn has been closely involved in the settlement of energy in the deregulated markets of California for the past ten years. Currently he oversees the complex re-settlement of energy transactions from prior years in various Federal Energy Regulatory Commission (FERC) proceedings.

Click here to enlarge image

Multidimensional modeling technology can vastly improve both the readability of complex models, as well as the ability to view the data in different perspectives. This example shows one of the simpler matrices developed by PowerMarket Consulting using multidimensional modeling technology Quantrix Modeler, displaying just some of the elements that are in the model.

Attributes and variables can have natural-language names (i.e. trading day), which make it very easy for anyone viewing the model to understand what formulae mean. The ‘write-once’ formulae for each output variable are shown in the lower windowpane of the matrix. Formula number 53, for example, displays the very last step in deriving the CAISO charge code 6011, using the exact variable names specified in the CAISO’s BPM.

Because attributes and variables are independent of the cell, they can be manipulated easily to view the data in a variety of formats. For example, Quantrix Modeler enables users of this model to drag an attribute around the data grid, which prompts the model to refresh in an entirely different perspective, all without reprogramming or developing individual worksheets. This example shows how output variables with four common attributes, corresponding to four small tiles on the top right corner of an output matrix, can be clearly represented in Quantrix Modeler.

Click here to enlarge image

This graphic shows a screen shot of settlement estimates from the Rozetta model developed by PowerMarket Consulting, using Quantrix Modeler. After the front office trades energy, it has to estimate how much it will be paid or charged.

When companies can leverage the capabilities of the settlements, and develop more accurate estimates, they can optimize their trading strategies to achieve their trading objectives. They also can avoid penalties and reduce incorrect charges and payments. These capabilities are very powerful, but often difficult, if not impossible to achieve using traditional spreadsheet technology.

While this graphic displays just one view of the summary data à‚— settlement estimates by charge codes à‚— it is easy to display the data in a more aggregated, for example by charge group or trade date by selecting the appropriate level from the drop-down menu.

No posts to display