Skip to content

Collecting Demand Data

mlamherr edited this page Aug 29, 2018 · 18 revisions

For the REISE simulations, each bus has a demand for electrical energy per hour. The buses are located in various geographic locations. In order to apportion the demand to these buses, the actual demand data has to be collected first.

Demand data generally comes from meter readings from Balancing Authorities (BA). A BA is responsible for maintaining the load, energy interchange, and energy generation balance for a specific area. Hence, a BA measures actual demand for electricity.

For hourly demand data from July 1, 2015 to present, the most complete and accessible data source is https://www.eia.gov/ as part of the EIA-930 survey. Data from each BA can be downloaded either by using an API call per BA or by bulk download or by downloading an Excel spreadsheet per BA. The API calls and bulk downloads are json files which can then be converted into tabular form where each row is an hourly time point in UTC time. These data will have missing hourly rows if there is no data collected for that hour. As of August 22, 2018, the Excel spreadsheets contain the most comprehensive demand data: the spreadsheets contain demand, demand forecast, net generation, and interchange data as well as data quality check results. For missing data, imputed values* are now available.

This document, https://www.eia.gov/realtime_grid/docs/userguide-knownissues.pdf, describes the data collection procedures, potential sources of error, and how data quality checking is performed. For simulation purposes, the largest source of error may come from the missing values, which EIA currently imputes. However, internal imputation methods can also be applied as our understanding of the data and the simulations improves.

Of the 38 BAs in the WECC system, 3 are in Canada or Mexico. Of the 35 in the US, 27 report demand while 8 do not.

Prior to 2015 (from 2006), hourly demand data are available in FERC (https://ferc.gov/docs-filing/forms/form-714/data.asp, info provided by support staff at WECC). These data are not likely 'cleaned' like the the EIA data; however, the EIA-930 data cleaning procedure is published and can be applied to the earlier data.

APPENDIX There are also some demand data in http://www.oasis.oati.com, for some BAs. Oasis is where BAs keep various data; not all BAs participate, as shown in Table 1 in the Appendix. The BA's with hourly or daily data (for some or all of 2010-12) include BPAT, EPE, IPCO,PGE, PSCO, PSEI, WAUW. Although limited, these data may be useful towards crosschecking extrapolations other data.

Table 1: Availability of 2010-12 demand data in OASIS

BA In oasis? Comments
AESO N
AVA N
AZPS http://www.oasis.oati.com/azps/ study files
BANC N
BCHA https://www.oasis.oati.com/bcha/ no data
BPAT https://www.oasis.oati.com/bpat/ hourly available
CFE N
CHPD N
CISO Y available
DEAA N
DOPD N
EPE https://www.oasis.oati.com/epe/ daily loads from 2012?
GCPD N
GRID N
GRIF N
GRMA N
GWA N
HGMA N
IID https://www.oasis.oati.com/iid/ no demand data in oasis, has water use forecast monthly in website
IPCO https://www.oasis.oati.com/ipco/ daily load 2010-12
LDWP https://www.oasis.oati.com/ldwp none in oasis
NEVP https://www.oasis.oati.com/nevp/ current demand only?
NWMT https://www.oasis.oati.com/nwmt/ has hourly load 2005-18, downloaded
PACE https://www.oasis.oati.com/ppw/ no data
PACW https://www.oasis.oati.com/ppw/ website has studies pre-2015
PGE https://www.oasis.oati.com/pge/ has demand file but not accessible?
PNM https://www.oasis.oati.com/pnm/ has reports, but not accessible ?
PSCO https://www.oasis.oati.com/psco/ has 2012-17 ‘minute’ “ACE” data – what is “ACE”?
PSEI https://www.oasis.oati.com/psei/ has demand data, 2010-11, hourly
SCL N
SRP https://www.oasis.oati.com/srp/ no data
TEPC N
TIDC N
TPWR N
WACM https://www.oasis.oati.com/wacm/ no data
WALC https://www.oasis.oati.com/WALC/ yearly peak transmission?
WAUW https://www.oasis.oati.com/wapa has hourly data pre-2015 but not yet clear what it is
WWA N

*Still awaiting documentation on how imputation is done.

Clone this wiki locally