-
Notifications
You must be signed in to change notification settings - Fork 14
2018 02 20 tag agenda
Date: February 20, 2018, 1pm-3pm PST (web conference)
Slides are posted at https://github.com/lbl-srg/obc/tree/master/meetings/2018-02-20-tag
Join from PC, Mac, Linux, iOS or Android: 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 available: https://lbnl.zoom.us/zoomconference?m=_h5BuJ686mPy3rWEyKf4NROxLaeOV5J8
- 1st case study and dissemination (start at 1pm)
- Guideline 36 release
- Next case study (1:10 pm)
- Chiller plant sequences (1:35)
- CDL export (2:00)
- Verification (2:10)
- Design tool (2:20)
- Timeline (2:35)
- Open discussion (2:40)
- Michael Wetter
- Paul Ehrlich
- Philip Haves
- Jianjun Hu
- Milica Grahovac
- Paul Switenski
- Brent Eubanks
- Jay Santos
- Mark Hydeman
- Marina Sofos
- Gerry Hamilton
- Janie Page
- Ira Goldschmidt
- Karen Perrin
- Dave Robin
- Charles Holleran
- Rick Stehmeyer
- David Prichard
- Amir Roth
- Jan Baker
Michael presented the main inputs, methodology and outcomes of the secondary sequence comparative case study. David Prichard asked about any industry interest in the CDL translation.
Michael described the desired setup of the case study and the necessary inputs. Suggesting chilled water plant coupled to building loads. Targets are to demonstrate that LBL team can translate CDL, generate English language translation and as a stretch goal go all the way to a product line.
Regarding potential chiller load data sources: Phil mentions Hamilton/Stanford - a possibility to extract loads on the central chilled water plant (building 2) - apply ASHRAE 1711 sequences; another option are the Oracle buildings with possibly good data. Rick mentioned that there may be some simple logics used in Kaiser Permanente projects. He also asked if there is any criterion on searching for the candidate case buildings. Rick says they are commissioning chiller plants for hospitals in Vermont. Phil introduces chiller sequence source (ASHRAE RP-1711). Paul on the selection of the building - expresses concern about the scope of the case study. Michael elaborates on feasible steps (use model in the loop, demonstrate CDL export in English language). Gerry says it’s a bit difficult to find a building that has 2 chillers, 2 boilers, 2 cooling towers. Most Stanford buildings have district cooling/heating.
Michael provides an intro and introduces sequences obtained from David Prichard and Steve Taylor. Jianjun explains how the LBL team interprets the sequence and then explains the intended implementation. TAG raises questions about the absence of the water side economizer and the lack of a condenser water side by-pass valve.
VFD issues (chiller, CHWP, CHP) - TAG suggests either all VFD or no VFD on the CW side. Arguments are raised that chillers and cooling towers should be VFD (due to federal/state/local codes). Mark suggests to stick to 2 identical chillers with a WSE and VFD and 2 boilers. It is usual to have identical chillers if the plant has 2 chillers. Discussion on pros and cons of both advanced/high performance sequences (demonstrate new capabilities) and conventional/solid sequence (just having them work properly). Ira: LBL’s sketch shows a typical chiller plant that is "slightly cutting edge".
Michael presents diagrams that explain the approaches to sequence translation and does a demo of how the English language documentation gets generated. TAG seems very interested and would like to review the current status. Rick raises some translation issues if the functional block implementations in different languages take different sets of inputs. Michael suggests a few solutions, such as replacing CDL say PID blocks with a Niagara PID block, although some compatibility issues may remain. Michael said the ultimate test will be whether the installed sequence passes the verification test relative to the control signal computed by the original unmodified specification. Alternatively, if a vendor wants to offer their control blocks, they could be put in a library and the designer could use them before exporting the sequence. Then it is the designer's choice to select a particular implementation. TAG raises validation concerns if replacing CDL blocks with proprietary ones in terms of loss of quality.
Michael presented the sequence verification concept.
Phil reports on Control Design Tool scoping. Paul presents a list of useful deliverables. Discussion about the unfamiliarity with some of the tools involved. Michael said that OpenStudio SDK and PyFMI are tools a user will not interact with directly.
Select a building that can provide real building performance data suitable for implementation of real chiller control sequences
Q.6. Demonstration with actual measured control response (need any input signals that go into and out of the controller, as well as controller parameters, such as control gains). In absence of such data, we would need an alternative simple control sequence, so that we can do a comparative study.