Hack Davis 2018 Data Overview

Version 1

    HackDavis Documentation provided by OSIsoft and UCDavis Energy & Data Group


    Business Case

    We have real time data from many meters on campus that measure building electricity, heating (steam/condensate), cooling (chilled water), gas, and domestic water usage. We also have meta data on buildings (square footage, building type, construction date)


    Business need: faster and better ways to use, visualize, and get value out of the data. This could be accomplished through being able to more quickly sort through usage data, having access to more novel or insightful visualizations, or having better metrics or analytics on the data.


    Example questions of interest:

    • How are the buildings doing right now? Yesterday? Last week? Last month? Last year?
    • Which buildings should we focus on for savings compared to other buildings?
    • Which buildings are performing the best/worst?
    • Which buildings are performing worse than they used to be? Do any have usage spikes?
    • Can better savings be achieved on nights or weekends?
    • How should the buildings be categorized? Are there several unique usage profiles?
    • Can we answer the above questions for only certain subsets of commodities, building types, or building sizes?

    We also would like to be able to download the results quickly for larger, more customized queries of the data. These are general guidelines and exemplary ideas. Feel free to think outside the box and come up with innovative ideas that bring more value out of the data.

    Introduction to CEFS (PI System Database to be used for HackDavis)

    CEFS stands for “Campus Energy Feedback System”. It is one of the databases that UC Davis uses to  view and organize their building utility meter data and as a back end for publicly available dashboards.
    CEFS AF structure and Attribute ListAll information is nestled under the “UCDAVIS” element as part of the CEFS AF database. Data for all buildings found on CEED can be found under the “Buildings” element. Under each building are the elements for the commodities with metered data. The attributes that contain the time-series data are described below. In general, the most useful attributes for data collection are: demand tags (source units or common units) and usage tags (cumulative counter, annual usage, and monthly usage). Note the following before finalizing your results:

    • Demand (instantaneous) data are separated by the source units for the commodity and the common units among all commodities.
      • Source unit demand attributes are named as “commodity_demand” (i.e. chilled_water_demand). Electricity is an exception; the name is “kW”
      • Common units are kBtu/h. These attributes are named as “Demand_kBtu”

    • Usage (over time) data can either be calculated using a cumulative usage tag or pulled at a specific time using a preset tag.
      • Cumulative usage tags contain “Counter” in the attribute name. Electricity is an exception; the name is “kWh”. THESE TAGS ARE IN SOURCE UNITS
        • Usage can be calculated over any time range by taking the cumulative use value at the desired end time and subtracting from it the desired start time. For example, (1/2/2016 00:00:00) - (1/1/2016 00:00:00) will yield the usage for January 1st, 2016.
      • Annual_Usage attributes are calculated as a rolling annual usage. For example, 1/2/2016 00:00:00 refers to the usage from 1/2/2015 00:00:00 to 1/2/2016 00:00:00. THESE TAGS ARE IN COMMON (kBTU) UNITS
      • Monthly_Usage is a monthly counter that resets on the first of each month. To retain the monthly usage using this tag, pull the data at the last day of the month at 11:59:59 PM. For example, 12/31/2016 23:59:59 represents the usage for the month of December, 2016. THESE TAGS ARE IN COMMON (kBTU) UNITS

    All relevant weather information is under the Element called Weather. Outside Air Temperature, Heating Degree Days and Cooling Degree Days can be found as attributes in this Element.
    An important metric that describes how a building is performing is the “Energy Use Intensity” (EUI). This quantity is the energy usage normalized by the building square footage. This quantity is calculated based on the rolling annual usage for each commodity and rolled up to the building level. There is also a cost estimate for each commodity that multiplies the rolling annual usage by the commodity rate for each commodity and then calculates a sum in the building-level element.
    There are several meta data items that may be useful for the hackathon. First, the square footage is provided in the AF element for each building (as well as in each meter element). There is also a table  in the database called “Building Data_New” that contains extensive information about the buildings  on campus. Key fields to look at are:

    • Constructed: date the building was constructed
    • Primary/secondary usage type/percent: the primary/secondary use of the building and the percent of the building square footage allocated to that use


    Login Information and web address

    Here is the login information to be used by students (from David Trombly): https://ucd-piwebapi.ou.ad3.ucdavis.edu/piwebapi/Login: ou\piapihack2018
    PW: “Go $ave energy, 2018!”

    PI Web API Example Queries


    1. Find all the Air Handling Units (AHUs) elements (assuming AHU is in the name, not through a template)
      1. https://{ServerName}/piwebapi/elements/search/

    2. Find all the the AHUs in the building Ghaunsi (assuming there is an attribute named “Building”, and for each one found list its attribute values.

      This problem requires multiple requests, one request to get all the elements to fit the search criteria, and another to find all the values for the elements attributes. This can be done in a batch POST call.

      1. The first batch subrequest should add onto problem 1 to add a attribute value filter in order to find the AHUs in the Ghaunsi building:

    }&query=Name:*AHU* “|Building”:=Ghaunsi

    1. The second sub request should contain a request template referencing the first subrequest, substituting the WebIds of all the returned elements from the first subrequest into the second requests streamsets/value call.

    The resulting body of the POST batch request (sent to https://{ServerName}/piwebapi/batch) will be:

      "firstSubRequest": { 
          "Method": "GET", 
          "Resource": "https://{ServerName}/piwebapi/elements/search/
        ?databaseWebId={WebID}&query=Name:*AHU* “|Building”:=Ghaunsi”
      "secondSubRequest": { 
           "Method": "GET", 
           "RequestTemplate": {
                "Resource": "https://{ServerName}/piwebapi/streamsets/{0}/value"
           "ParentIds": ["firstSubRequest"], 
           "Parameters": ["$.firstSubRequest.Content.Items[*].WebId"] 

    1. Find all the tag names on the PI Server containing *Demand*kbtu*
      1. https://{ServerName}/piwebapi/dataservers/{DataServerWebID}

    2. Find all the attributes that are in the element template: “Electricity_Ceed” and start with the letter ‘A’
      1. https://{ServerName}/piwebapi/attributes/search/?databaseWebId={WebId}
        &query=Element:{Template: “Electricity_Ceed”} Name:=A*

    List of PI Templates in CEFS Database

    Link to the Current Energy Dashboardhttp://ceed.ucdavis.edu/#!/

    Support Articles

    https://techsupport.osisoft.com/Documentation/PI-Web-API/help/controllers/search/actions/quer y.html

    List of Buildings


    Primary Use: LAB

    Primary Use: Classroom

    Primary Use: Office

    Primary Use: Community
    • Center for Comparative Medicine
    • Hoffman EAPL
    • Bainer
    • Meyer
    • GBSF Genome
    • Chemistry Annex
    • WHNRC
    • VM3A
    • Robert Mondavi Institute
    • Robbins Hall
    • Plant and Environment Sciences
    • Ghausi
    • Briggs
    • SciLAB
    • EPS
    • Primate
    • Wickson
    • VM3B
    • Physics
    • Hutchison
    • RMI Brewery
    • Young
    • Valley Hall
    • Giedt Hall
    • Student Health and Wellness
    • TAPS
    • Maths
    • MRAK
    • FPS Trincero
    • Rifle Range
    • SCC
    • ARC
    • Tercero Dining Commons
    • Tercero 1
    • Segundo Regan
    • Segundo North
    • MU
    • Tercero 2
    • Hutchison Child Dev
    • Tercero 3