Agent-Based Microsimulation Models

Jackson Associates CEDMS and REDMS end-use energy demand models, which were first applied in 1982, were extended in the late 1990's to incorporate MAISY® Utility Customer Databases information and to reflect current utility, retail energy provider, equipment manufacturer and regulator forecasting and analysis needs.

The extended models are part of the MAISY Agent-Based Microsimulation Forecasting System and are referred to as MAISY Residential, Commercial and Industrial MFS models.

Agent-based microsimulation forecasting models have great conceptual appeal. A statistically representative sample of residential, commercial or industrial customers for any geographic area is developed from the MAISY Utility Customer Databases. Information on building, equipment, operating hours, end-use energy use and other data (including option hourly load data) are available for each customer. Current year energy use can be calculated by applying weights to each of the sample customers and summing across all customers in the sample.

Energy, hourly load and equipment forecasts for future years are provided by updating the sample of customers for the first forecast year. A sample of new utility customers is added to the process to reflect new construction; customer weights in the existing sample are adjusted to reflect demolitions of existing buildings. The new customer sample reflects recently new customers drawn from the MAISY databases. The same process is completed for each year in the forecast period.

Energy use and equipment characteristics of each sample customer can change in each forecast year. Existing equipment wears out and is replaced. The efficiency and energy use for these end uses is changed to reflect new equipment. For new construction, efficiency and energy use and fuel choices are incorporated in individual sample customer records. Energy price changes cause changes in utilization of most end-use equipment (e.g., increasing natural gas prices result in thermostat changes). Utilization and fuel choices are modeled with econometrically derived parameters and efficiency changes are modeled with econometrically derived parameters and with information on efficiency possibilities and appliance and building standards.

Customized Applications Provide Flexibility

MAISY MFS models begin with a comprehensive, basic characterization of energy use at the customer level. This comprehensive representation allows the models to reflect nearly any issue of interest to model users. Specific equipment sales, 8760 hourly load profiles, end-use energy use and more traditional energy forecasts can all be provided within the MAISY MFS model framework.

Jackson associates works with clients to identify specific forecasting and analysis needs and provides customized models to meet these needs.

MAISY MFS Advantages

The MAISY agent-based microsimulation forecasting models provide a number of advantages compared to other approaches. The system:

Uses a statistically-representative sample of customers in a transparent way to determine energy use at any desired level of aggregation
Applies technology and customer-detailed analysis of building and energy-using systems in a process that simulates choices actually made by individual customers. This process also avoids "double counting" problems common with other approaches.
Explicitly represents technology-detailed impacts of DSM measures
Uses stratified customer samples to support the evaluation of user-specified customer segments and to accomplish vintaging of buildings, equipment and program impacts
Incorporates modeling and customer analysis methodologies developed in DSM, integrated resource planning, forecasting, market and technology analysis applications over the last twenty years
Applies a scalable structure which can easily be modified or extended to incorporate new analysis requirements
Provides an extension of widely-accepted analysis methodologies used by electric and gas utilities, states, regional organizations and federal government and other organizations
Provides customized applications to meet specific needs of clients
Takes advantage of utility customer data available in the widely-used MAISY utility customer databases
Provides more targeted information at less cost than alternative approaches

Building and End-Use Energy Detail

Building and end-use detail is also customized to meet specific client needs. Typical detail is shown in the following table.

Model Structure Summary

The concept of microsimulation is intuitively appealing in that results are based on technology-level analysis performed on individual customers comprising a statistically representative sample of all utility customers. Analysis results are statistically expanded to the population of customers to develop an accurate estimate of energy use and hourly loads. This process is similar to that used in surveys where responses from a sample of customers are used to develop reliable estimates of responses that would be provided if all customers were surveyed.

The MFS sample of individual utility customers is drawn from the MAISY Utility Customer Databases which have been developed with more than one million residential, commercial and industrial utility customers throughout the US and Canada.

In each year of a MAISY MFS forecast period:

Some customers are removed from the sample reflecting the demolition of buildings and customers who leave the service area.
New customers are added to the sample to reflect service area growth in the customer population.
Each customer's equipment holdings, building thermal characteristics and equipment operation are modified, as required, to reflect end-use equipment replacement, new equipment purchase, building shell upgrades and price-induced changes in equipment operation.
Each customer's 8760 hourly loads are determined by summing across end-uses where end-use loads reflect changes identified in the previous item.
A robust sample of customers is used in the MFS to support analysis and forecasts at any level of detail from technology-specific to total system results. Customer samples are stratified by psychographic and firmographic variables (including building vintage) as well as by rate class and climate zone. Sample weights associated with each sample customer will, when applied to customer characteristics, provide accurate estimates of number of customers, customer building, equipment and operating characteristics, energy use, and 8760 hourly electric loads.

When run without any DSM, demand response, technology product or other inputs, the MFS provides baseline energy use forecasts reflecting market-driven changes in equipment efficiency, equipment and fuel choice and equipment utilization. Model runs with these non-baseline inputs modify this process to reflect the impacts of these programs and other activities.

The MFS can also be applied to evaluate traditional utility DSM costs and benefits. Replacing existing equipment or building thermal measures with those that are technically feasible reflects technical potential. An economic potential run makes changes in individual customer equipment and building characteristics that are economically justified, and so on.

MFS models provide energy use forecasts at detailed end-use, building and sector (residential, commercial, industrial) level. The models can also provide hourly load forecasts that range from peak demand to full 8,760 hourly loads. Additional energy, building and equipment detail (e.g., installations of small combined heat and power systems) can also be provided.

As schematic of the MAISY MFS is shown below:

MAISY MFS System Schematic

Application Areas

The following application areas have been addressed with the MAISY Microsimulation Forecasting System (MFS)

Service area electricity and natural gas forecasts including peak demand and hourly loads
Costs and benefits of alternative DSM and demand response programs
Individual DSM program design
Detailed rate structure analysis
DG and combined heat and power market penetration and impact analysis