Skip to content

2018 07 12 team meeting agenda

Jump to bottom
Michael Wetter edited this page Jul 13, 2018 · 4 revisions

Agenda for team meeting

Date: July 12, 2018, 10:00-11:30 Pacific Time

Dial-in information: https://lbnl.zoom.us/my/mwetter

Or iPhone one-tap (US Toll): +14086380968,6614042296# or +16465588656,6614042296#

Or Telephone:

Dial: +1 408 638 0968 (US Toll) or +1 646 558 8656 (US Toll)
Meeting ID: 661 404 2296
International numbers: https://lbnl.zoom.us/zoomconference?m=_h5BuJ686mPy3rWEyKf4NROxLaeOV5J8

Or an H.323/SIP room system:

H.323:
    162.255.37.11 (US West)
    162.255.36.11 (US East)
    221.122.88.195 (China)
    115.114.131.7 (India)
    213.19.144.110 (EMEA)
Meeting ID: 661 404 2296

SIP: [email protected]

Agenda

  1. upcoming deadlines
  2. implementation of control sequences for primary systems
  3. CDL export to English language and control product line
  4. verification of control sequences
  5. case study update
  6. next steps

Minutes

Participants

  1. David Pritchard, Arup
  2. Paul Switenki, Arup
  3. Gerry Hamilton
  4. Mark Hydeman, Google
  5. Rich Rockwood, Oracle
  6. Lyn Gomes, kW Engineering
  7. Janie Page
  8. Michael Wetter
  9. Jianjun He
  10. Milica Grahovac

Presentation

Michael goes through the presentation.

  1. Upcoming deadlines
  • Q7 deadlines discussed
  • NREL coordination with Open Studio
  • Case study report due in 2 months, now in progress
  1. Implementation of control sequences for primary systems
  • Implemented sequences according to ASHRAE RP1711
  • Status: sequences need review and validation in open & closed loop; LBNL team will reach out to broader OBC team as needed
  1. CDL export to English language and control product line
  • Developing export of textual and graphical elements in HTML and JSON formats
  • Support for graphical annotation ongoing – block diagram, visual representation (Modelica format export in form that PDF or HTML can use consistently)
  1. Verification of control sequences
  • (chapter 9 of draft report shown) – explaining the purpose, structure and an example of a verification test.
  • Additional change suggested by Dave Pritchard to be able to confirm economizer operation.
  • Read into Modelica model, unit conversions as needed (to SI).
  • 9.3.3 time series work in progress
  • 9.2 figure shown as example of data flow to setpoint computation block
  • Example validation test: HTML file from simulation vs. actual measured data – near perfect agreement! Trend correlation within 5% (continuous vs discrete time may be source of distinction).
  • Test duration: 90 hours (3+ days)
  • Discussion on the scope of the validation tests - LBNL team can verify the control sequence and not any response of the physical equipment, unless recorded as a trend that is used as an input/output to such a sequence. Allows commissioning work to confirm approach (also can be used during design work). Goal is to confirm control is accurately represented. Confirm connections made correctly; implementation is as specified.
  1. Case study update
  • Similar validation test set up planned for Oracle case study with the primary system control sequences.
  1. next steps

Team discussion

Difficulty is judging how well data matches. New code from team: Band around time series calculated then look at new time series band to see if it stays within tolerance set. If signal switching gets delayed, the tool will capture it. Looking to release it as open source soon. Will this show problem in communication system? No, that is separate from this. This shows what goes in/comes out of control sequence. What about time delays built into system (e.g. numerous 1-minute blips)? Can system sort these out so false positives can be minimized? Yes – there are tight tolerances – on time sequence and on required accuracy. Can be set as needed to catch these, based on system response times. Guidance on time and value of tolerances to be developed in future using sample data from real operations.

Case study discussion: Tour end of June by Oracle/SunBelt controls. Trending data from low to high load (to both chillers in use). Trend data now being analyzed as are control sequences. Have 1-minute interval trend data for 12 hours, several stages recorded based on one economizer, two chillers. Chiller staging does not yet include water side economizer staging but working on this so can be used in future case study. Testing open and closed cycles. Water side economizer sequence exists in real buildings, but information sent to team did not include this. Latest 1711 sequence would be of interest to team, if available. Actual sequence implemented is needed for case study, but wider need is to populate library with sequences used in real life (guideline 36 sequences). Differences between 1711 draft and Oracle not huge but knowing when to implement economizer is key, based on measurable data (wet and dry bulb temps). This is more detailed and is in development. This is of interest to team when available. Can we identify building in public reports and release graphics of system? Would be helpful to know soon how much of information we have can be put into the report. Rich/Oracle notes this will require some research – expect answer by end of day tomorrow. Data files as Excel vs CSV? Not a big deal; easily converted. Status: Just over one year to go. Behind on control sequences for primary systems. Need code review and more detailed economizer model. Also case study is still on target to complete by end of August, but this is tight. MW on vacation beginning next week, but work can continue. Market transformation picked up by ARUP; expect report soon for DOE review. Staffing down but working to hire new people.

Questions: Fault diagnosis in code? Not yet. Discussed with NIST – planned to be added to Guideline 36, but not yet included because not clear how much detail/tuning required from specific buildings. Since not yet in Guideline 36 for VAV boxes; air handler has fault detection. Not sure if sufficient resources in project to implement this. Is sequence tuned properly? Lack thresholds to apply. (Is equipment short cycling, is hunting going on?) Tests can be codified if needed to test, based on guidance received. Loop tuning [continuous monitoring] is planned to be implemented in Guideline 36 but not yet there. Once nailed down, can simulate. MW will talk with PI on that project. Would be useful to confirm stability of system prior to additional studies, determine whether fits into scope of this project. Audience noted that during real time commissioning, don’t always have full control of all systems so simulations may not match data because of transients associated with start/stop of systems. MW notes that the focus is on what goes in/comes out of controller. Valves could look like hunting but really adapting to changes in supply air temperature, for example. Audience noted that could set limit based on stage of process (e.g. “goodness factor” of, for example, 80% when bringing system up) to allow for temporary instability from known source.

Meeting adjourned at 11am. Next meeting late August or early September.

Next Steps

  • Waterside economizer sequence (1711/Oracle building) will be sent to team
  • Team to reach out about any chiller control sequences implementation issues
  • Oracle to notify team about permissions to publish.
Clone this wiki locally