ZIP Area Utility Customer Energy and Hourly Load Databases

ACS - MAISY Residential Socio-Economic, Building,Commuting, Energy Use and Hourly Loads ZIP-Detailed Databases

ACS-MAISY Database Development Detail

ACS-MAISY Databases use actual individual household socio-economic data provided by US Census American Community Survey as a platform to extend each household's information to include a variety of other important data items. These merged data items include information for individual households in dozens of supporting databases including appliance holdings, energy use, metered hourly loads, travel, building, geodata, and emissions. The data merging process is an AI non-parametric k-nearest neighbor (KNN) machine-learning analysis with regression refinements providing superior accuracy and reliability for the combined household database. More than 6 million individual household records are provided in the ACS-MAISY Databases.
...
ACS-MAISY Databases Use AI to integrate ACS and dozens of household databases

ACS-MAISY Databases Use AI to integrate ACS and dozens of household databases

ACS-MAISY Database AI Methodology

Database records from supporting databases are identified as belonging to each ACS household's "neighborhood" with a weighted distance metric that reflects several dozens of the household’s ACS characteristics. Variables common to both the ACS household and each household in the supporting databases are used to calculate the "distance" between the ACS and each household in the supporting databases.

The KNN algorthim is referred to as a machine-learning system because weights on each variable in the distance measure can be determined with repeated evaluations of sample data to identify the most accurate distance measure with respect to the particular merged variable.

Data from the supporting database households with the smallest distance measure is merged with the ACS household record. Any difference between the ACS household and the closest neighbor variable values is adjusted with variable-specific statistical regression-based data refinements.

This AI approach extends ACS individual household records to include dwelling unit physical characteristics, appliance holdings and energy use and costs, GHG emissions, ZIP code location, commuting miles, and end-use/whole building hourly electric kW loads.

The non-parametric (read flexible) AI KNN estimation process is far superior to traditional engineering and other fixed parameter energy use estimating models like those used to develop DOE and NREL energy use and hourly load estimates. The AI KNN application applies actual data derived from similar "neighborhood" households whereas the DOE/NREL approach relies on models that reflect fixed relationships applied to all households, missing nuances that exist in small geographic areas and detailed household/dwelling unit segments - that is data items that are not explicitly reflected in the inflexible DOE/NREL models.

The schematic illustrates the ACS-MAISY AI KNN process. Individual ACS household records are first geo-located to identify ZIP codes. ACS household record matching is accomplished with the KNN process described above to integrate energy use, appliances, hourly loads, dwelling unit, emissions, commuting data, climate, place, county, metro and utility names, and other information and then provided as a comprehensive household database.

ACS-MAISY Increases Geographic Granularity

ACS-MAISY ZIP-level Databases provide 11 times more granular geographic detail than ACS public use databases provided by the Census Bureau. ACS public use databases identify each household location within a Public Use Micro Data areas (PUMAs) whereas ACS-MAISY household records identify more detailed ZIP code tabulation areas within each PUMA. An average PUMA includes about 11 ZIP code areas. Significant variation in household/dwelling unit characteristics typically occurs across ZIP code areas within each PUMA. Average ZIP code tabulation areas provide information on approximately 5,000 households whereas PUMA areas reflect aggregate characterizations of much wider geographic areas including approximately 56,000 households.

NEW: ZIP-level Detail and Crosstabs

In addition to more detailed geographic information ACS-MAISY Databases support ZIP-level crosstabs whereas ACS public use databases can only provide more geographically aggregated PUMA crosstabs.

Socio-economic, dwelling unit and other household characteristics typically vary widely within each puma. Both the increased geographic detail and ZIP level cross-tab ability make the ACS-MAISY Databases far superior at supporting detailed socio-economic, marketing, market sizing and other applications compared to Census PUMA-based analysis.

Data Formats Like You Want

Database information can be provided in spreadsheets with data for individual household records in separate rows, and/or processed by us to provide tables, cross tabs or any client-specified analysis results.

ACS-MAISY Database Items

The following table shows data item detail available with the ACS-MAISY Database. Clients can select any data groups or data items for any geographic area (single ZIP, county, metro, etc.) and data format/analysis preference including full CVS databases with each household data items on a single row, individual relational databases that can be merged (joined), tables, cross-tabs, etc.

Electric loads (kW) are available as 8760, 35040 15-minute, or monthly averages (weekday, weekend day and peak day). Load detail includes whole building loads and individual end use loads.

Not sure exactly what data items you need, or which data items are best for the project at hand? - Let us help! – just e-mail us with your questions and/or suggest a time to discuss. We provide free consultations to help identify the most useful data/analysis for your application. We also provide free telephone support to assist in client data applications after data delivery.

ACS-MAISY Residential Household Database Variables

ACS Data Items Household Record Id State FIPS number Household Weight At home (calculated variable) Dwelling Unit Type Lot Size Number of Bedrooms Heating Fuel Total Dwelling Unit Rooms Tenure Property Value Property Value Number of Vehicles Construction Year Householder Age Householder Hispanic Householder Race Household Income Children Age Categories Households with Limited English Number of Children Household Members Owner Monthly Cost # Household Person Records HOUSEHOLD MEMBERS Person Age Person Sex Time to Work Transportation Mode Marital Status Educational Achievement Time Arrived at Work Time Departed From Work	BUILDING DATA Dwelling Unit Total Square Feet Dwelling Unit Heated Square Feet Dwelling Unit Cooled Square Feet Number of Bathrooms Heating Equipment Detail Air Conditioning Equipment Detail LOCATION-RELATED DATA ZIP Code Tabulation Area 30-year Avg Heating Degree Days 30-year Avg Cooling Degree Days ZIP Code Name County Name Place Name Metro Area Name Utility EIA Name EMISSIONS Electric CO2e Emissions Natural Gas CO2e Emissions Fuel Oil CO2e Emissions LPG CO2e Emissions Total Household CO2e Emissions ANNUAL ENERGY COST Total Energy Cost Electric Cost Natural Gas Cost Propane Cost Fuel Oil Cost ANNUAL ENERGY USE Annual kWh Use (kWh) Annual Natural Gas Use (kBtu) Annual Propane Use (kBtu) Annual Guel Oil Use (kBtu)	END-USE ENERGY USE Electric Air Conditioner Electric Cooking Electric Dryer Electric Dishwasher Freezer Indoor Lights Electric Miscellaneous Electric Pool Pumps Refrigerator Electric Space Heating TVs Electric Water Heating Fuel Oil Other Fuel Oil Space Heating Fuel Oil Water Heating Propane Other Propane Space Heating Propane Water Heating Natural Gas Other Natural Gas Space Heating Natural Gas Water Heating HOURLY LOADS 8760 Hourly & 15-min Whole Building & End Ese Format is as follows: m1-d1-hr1 m1-d1-hr2 m1-d1-hr3 . . . . . . . . . . m12-d31-hr22 m12-d31-hr23 m12-d31-hr24

Additional MAISY Residential Database Background Detail

Click here to see a variety of related background information.

Save Cost, Time and Hassle and Improve Analysis Accuracy With ACS-MAISY Databases

Whether you want to extend socio-economic evaluations to the ZIP level, expand household information with new data items (e.g. energy bills, emissions, floor space), drill down on specific ZIP household segments or are looking to improve you market analyis and target marketing, ACS-MAISY Databases provide new and more accurate analysis results.

Some applications:

Identify variations in ZIP-level socio-economic variable combinations to focus social welfare programs (e.g. percent of households in income/#children/workforce participation/householder age categories).
Evaluate variations in ZIP-level socio-economic variable combinations such as energy costs by income segments and household characteristics, dwelling unit owner monthly cost by income,householder age, etc.
Evaluate market share across ZIP code areas by combinations of household and other variables to inform future expansion and marketing strategies.
Identify ZIP codes with specific targeted income/demographic/dwelling unit/householder education etc., characteristics for program or marketing efforts.
Assess likely hourly load impacts of EV charging on local distribution grids.
Identify ZIP areas with large peak hourly load reduction potential and household characteristics associated with demand response program participation.
Identify high or low load factors for specific technology applications.
Evaluate ZIP-area weather risk (% sales in space heat, AC).
Correlate loan write-offs and combined household characteristics to revise deposit requirements.
Evaluate market characteristics in underperforming ZIP areas to revise messaging and sales activities.
Identify high growth markets based on combined household variables.
Distinguish between stable and transient ZIP markets.
Identify ZIP areas that can be targeted for upgraded products and services based on dwelling unit age, appliance holdings, householder age and other factors.
Apply ZIP-detailed customer acquisition and retention data to refine pricing models.
Conduct other socio-economic, market and sales analysis.

Database Costs

ACS-MAISY Databases include proprietary information belonging to Jackson Associates and are licensed to individual users who agree to use the information only within their own organization. Information is provided in any of the following formats: (1) databases with individual household records in CSV files for client-specified database items and ZIP codes, (2) client-specified tables/cross-tabs for any combination of database items and ZIP codes. Database costs depend on data items and ZIP codes included in the data request. Data extraction costs start at $195. We provide free consultations to discuss data needs and will be happy to suggest appropriate data selections for you project.

We understand the difficulties in identifying exactly the information you need for your project. Casting a wide net costs more; however, trying to minimize data collection costs can dramatically reduce information value. We have worked with start-ups and some of the largest energy-industry organizations to thread this needle based on our energy data analysis experience. We can help identify options and suggest the most effective data develoment strategies for your projects. – just e-mail us with your questions and/or suggest a time to discuss. We provide free consultations to help identify the most useful data/analysis for your application. We also provide free telephone support to assist in client data applications after data delivery.