Description:  15 minute average power data (KW) for ADR participants on Trial Days.  

Automated Demand Response trials within the NTVV project installed Building Management Systems within 27 commercial buildings within the Bracknell area.  
The project was then able to enact load shed events on the building with differing notice periods.

The data are stored in 4 subdirectories corresponding to the four trial seasons.The data are in 99 csv files split between the four trial seasons. Each CSV file contains 15 minute average power data for one participant over the load shed events in one season.

The trial seasons are:
Winter (30/11/2015 - 26/2/2016),
Spring (28/2/2016 - 25/5/2016),
Summer (31/5/2016 - 26/8/2016),
Autumn (30/8/2016 - 25/11/2016).

In each season subdirectory there is one excel spreadsheet containing the schedule of load shed events and the participant groupings to which load shed events were applied. There are also pdf and csv format copies of the event schedule and participant groupings.

The following are the different types of load shed that were planned:
16:00 System Peak - Fixed at 16:00
Local System Peak - Analysis carried out on feeder of building to determine the best time for a load shed event for that area of the network
No notice events - events lasting up to 1 hour were scheduled with 2 minutes notice

Events up to 4 hours in length were scheduled.  Please look at the schedule spreadsheet in each season to see the spread of load-shed styles. 

A buildings load shed size was calculated through comparison of day before/ week before / and baseline analysis.

There are four types of buildings that have been tested:-

Data Centre
Educational Centre
Leisure Centre
Local Office
MNC (Multi-National Corporation)


The format of each csv file is as follows:

For each load shed event there are three lines of header information followed by lines of 15 minute data. 

Different events are separated by 2 empty, comma separated lines:,,,,

Header description:

Header line 1: Times,Event<event number>,,,
Header line 2: ,1 week prior,1 working day later,Baseload,Day of shed
Header line 3: ,<date of 1 week prior data>,<date of 1 working day later>,,<date of load shed event>

Data description: 

All lines: Column A,Column B,Column C,Column D,Column E

Column A: Time (HH:MM)
Column B: 1 week prior - the test buildings 15 minute average power consumption one week before the shed day (kW)
Column C: 1 working day later - the test buildings 15 minute average power consumption one working day after the shed day (kW)
Column D: Baseload - the test buildings baseload 15 minute power consumption data, calculated as a rolling 10 day average of the buildings 15 minute power consumption at the same time of day.  Note that this is not included for any Saturday shed events.(kW)
Column E: Day of Shed - the test buildings 15 minute power consumption on the shed day. (kW)


Note that if any of the comparison days fell on a day which also had a shed event planned, a different comparison day was picked.  It is therefore possible to see '1 week prior' or '1 working day later' dates that are not as expected.

Background information on the NTVV project and the ADR trial can be found in report http://data.ukedc.rl.ac.uk/simplebrowse/edc/Electricity/NTVV/Reports/SDRC_9.1c_Demand_Side_Response_Evidence_Report_1.0.pdf and at www.thamesvalleyvision.co.uk