Building II has a trane 40-70 ton CGAM model air conditioning unit.

Monitoring the AC unit

Techview

    The trane ac has a small computer with human interface at the ac unit. It also has the capability for remote monitoring  via an rs232 cable and a trane software called "TechView"

• Techview runs on windows (I'm using windows 7)
• downloading techview:
• google trace techview download.
• You need to download tv14.0 and then tv14.0 sp6
• When you click on download, they send you an email with a link to download (one time only)
• This is a hassle if your mail is not on windows.. just download the file to a thumbdrive.
• running techview:
• install rs232 cable from hmi on ac to your laptop (via an rs232 to usb converter)
• start techview
• i got com3 to work
• To enable logging,
  
• options..
  
• sampling time.. i set it to 2 seconds
• enable logging
• reports:
• The output datafiles:
• there are two files generated during logging. both are ascii files (although they use all 8 bits)
  
• xxxx.log .. generate log file
• xxxx.txt .. log file converted to txt. I used this for monitoring
• the records in the file are delimited by cr/lf..
• fields are separated by htab (horizontal tab.. binary 9)
• The  xx.txt file
• the first line as 450 ..
  
• the 2nd line has the field names
• there are 274 fields, separated by tabs
• the rest of the file is data.
• be careful, there are adjacent tabs for some empty fields.. don't combine adjacent tabs into 1 field
• I first remove all ascii codes above 127 (Eg.  deg symbol, etc.) since cutting and pasting doesn't like them..
  
• tr -d '\200-\377' < xxx.txt   > dataReport_yyyymmdd_hhmmss.txtC
• I'm currently storing them on linux at; /share/megs/phil/x101/ac/bld2/trane/logs/yymmdd/

Reading the data into idl

• I've written a number of idl routines to input and plot the trane data
• The routines are in: /share/megs/phil/x101/ac/bld2/
• trinplog: input the   dataReport_yyyymmdd_hhmmss.txtC  files.
  
• it returns a trI struct that has some fields that i selected out of the 274 fields:
• Abbreviations used in field names:
• evp: evaporater: where the heat is exchanged.
  
• chil: chiller: another name for the evaporator
  
• ckt1: circuit 1
  
• ckt2: circuit 2
• cmp1a,cmp1b,cmp2a,cmp2B: compressorXY
• EXV: expansion valve (at input to evaporator)
• SUCT: suction.. input at a compressor
• DISCHARGE: at  output of compressor
• DIF: differential
• PRES: pressure
• REFR: refrigerant
• EST: estimated
  
• suf:
• PC : percent
• T : temperature
• UNF: unfiltered
  

• field1	subfield	values	Notes
  TM		1461313897	secs 1970
  DAYNO		113.35529	of year
  MODE
  CKT1
  	EXV_CMDPC	0.0	percent
  	EXV_FLOWCMDPC	43.6	degF
  	AIR_FLOWPC	0.0	percent
  	AIR_FLOWCURPC	0.0	percent
  	CMP_INVOLUTE_PRES_DIF	0.0
  	DISCHARGE_EST_T	0.0	degF
  	DISCHARGE_PRES	220.4	psi
  	DISCHARGE_SAT_REFR_T	75.9	degF
  	DIF_REFR_PRES	108.7
  	EVP_INP_T	0.00	degF
  	SUCT_SUPER_HEAT_T	16.1	degF
  	SUCT_PRES	111.
  	SUCT_SAT_REFR_T	37.1	degF
  	SUCT_T	53.2	degF
  CKT2
  	same as ckt1
  CHILL
  	CHIL_RUN_HRS	24090.5	hours
  	AIR_T	75.4	degF
  	EVP_PUMP_CMD	1
  	EVP_PUMP	1
  	EVP_WATER_FLOW_SW	1
  	EVP_WATER_FLOW_ST_UNF	1
  	EVP_INP_WATER_T	49.2	degF
  	EVP_OUT_WATER_T	43.6	degF
  CMP1A
  CMP1B
  	same as cmp1A
  CMP2A
  	same as cmp1A
  CMP2B
  	same as cmp1A

  
  

• Adding elements to the struct:
• The struct is built inside trinplog:
  
• I create a string array holding: name :Type,$;searchstring    for each sub structure
• trmkstruct(): takes the string array and  creates a struct.
• the types are:
• F  float,
  
• I long,
  
• S string,
•  HM: hour:min.. returns float,
  
• SNO  string that starts with NO .. returns 0. else 1
• SEN: string starts with ENABLE .. returns 1, else 0
• SON: string starts with ON          .. returns 1, else 0
• trldstruct() is the used to parse the data into each record.

Plotting the data:

How to make the plots

• trmkplots(tri,yymmdd,....) will plot the data in a trI struct.
  
• doit.pro  : is a script that plots multiple days (on separate outputs). It is an example of how things work..
• It is in /share/megs/phil/ac/bld2/trane/plots/
  

Plot descriptions:

• Page 1: temperatures:
• Top: evaporator water temperatures, input:red (hot) output green (cold).
  
• I've put + at each sample of the red hot data.This shows when data was actually sampled
• 2ndFrame: differential water temp (hot - colde)
• The blue line shows when there is water flow.
• The red line  shows the water flow loss count. After each water flow loss, the system restarts. When the count gets to 4, a water flow loss causes a shutdown (the user has to reset diags to bring it back).
  
• 3rd frame: evaporator input refrigerant temp
• black is circuit 1, red is circuit 2
• The system switches circuits on and off. when the temp goes to 0, the circuit is not in use.
• 4th frame: suction saturated refrigerant temperature..
• This is output from the evaporator, input to the compressor
• bottom: discharge saturated refrigerant temperature
• this is output from the compressor (on the way to the condenser).
• Page 2: pressures, expansion valve
• top: suction pressure:
• input to a compressor. Black is circuit 1, red is circuit 2
• I've put + at the bottom of this plot showing when the circuit was on (in use) and off.
• 2nd: discharge pressure:
• output from a compressor
  
• I've put + on circuit 1 to show where we actually have data samples.
• 3rd: differential refrigerant temperature
• compressor output - input pressures.
• bottom: expansion value % open
• this is between the condenser and the evaporator. It throttles the amount of refrigerant going into the evaporator.
• Blowup plots about any water flow losses...
  
• i put a vertical blue line where it occurred.. 
• I redo the plots at  +/- A few minutes about each flow loss
• The times are usually those reported in the diagnostic log.. 
• Missing data:
  
• after  a power dip, the system would occasionally stop recording data.
  
• Looks like the connection was getting lost..
• So many days there are gaps in the data (i wasn't getting up to reset the connection after each power dip).

Daily plots:

• 160426: before this date, my windows laptop was set to go to sleep after 30 minutes (even when plugged in!). So the gaps in the data are from windows going to sleep...
  

• 160411:    11:04:16 to 12:15:07 (.ps) (.pdf)
• 1 second sampling
  
• 1 water flow  failure around 12.12 hours (12:07)
• 160419: 14:30:05 to 18:02:32 (.ps) (.pdf)
• 2 second sampling
• no data (15.2 to 16.05)
• two water flow failures: 15:30 (no data) and 16.59  (16:35)
• 160421: 9:30 to 17:50 (.ps) (.pdf)
• 2 second sampling
• 3 water flow failures:12:04 (nodata) (12:38, recorded), 17:47 (no data)
• there were lots of power dips (Eg 14:34)..
• I blew up each set of data (even if no failure)
• 160422: 09:30 to 17:17 (.ps) (.pdf)
• 2 second sampling
• water flow loss failures: 9:01(recorded),10:33,13,00(recorded),13:01(recorded),17:17 
• Lots of missing data... power dips
• 160423: 09:30 to 14:11  (.ps) (.pdf)
• 2 second sampling
• No water flow failures, but lots of power dips
• only have data 10:12 to 10:43
• 160425: 08:00 to 14:09  (.ps)  (.pdf)
• 2 second sampling
• no water flow failures
• 4 separate sections of data (lots of data dropouts)
• 1s 2 pages is all data, then i blew up each section separately
• 160426:  08:30 to 16:06 (.ps)   (.pdf)
• 2 sec sampling
• the 1st 2 pages show the entire day. I then blew up the 5 sections of data (power outages and water flow loss).
• 
  
• 160426: blowup of 4 water flow failures (.ps) (.pdf)
• 13:45 to 15:24.. this is a blowup of the 4 contiguous water flow failures
• You can see the fail count start at 0 and then count up to 4. When it reaches 4 (15.2 hrs)  the system stops doing auto restarts.
• 160427:  08:30 to 15:55 (.ps)  (.pdf)
  
• the water flow loss count remained at 4  for the entire day...
  

• directory: /share/megs/phil/x101/ac/bld2/trane
• techview diagnostics:
• ./diags/  .. i do a screen grab of techview diagnostic messages and put them here
• ./logs/Readme: i keep track of the data taken
• ./logs/yymmdd/  where i store the .txt files for each days data
• ./plots/   where the daily plots are located
• To run idl from the ./plots directory:
• idl71
• @../pro/trinit.pro
• then edit, run doit.pro
• ./index.doc
• lists the fields by topic
• n        m Description:
• n - counts the entries in this subtopic
• m - field index for this data (cnt from 0)
• Description: field description (after stripping off the topic (eg chiller:)
  

Documentation

• TRANE Installation, operation and Maintenance manual (.pdf)
  
• Building 2 A/C schematics (.pdf) (2013)
• Note a better (ie. not scanned by me) copy of the schematics starts on page 183 of the Installation,operation, and Maintenance manual.

 The Water pumping system (11aug16)

    The water pumping system was installed by a separate company. A single pump moves the water:

• output of pump
• to bypass valve
• this can route the water around the evaporator
• into the evaporator
• out of evaporator
• bypass will bypass to  here...
  
• down to individual units in rooms
• return to pump

• The system is controlled by a small honeywell computer.
• There is a differential pressure sensor between the input and output of the evaporator. 
• the honeywell cpu tries to keep 8 psi pressure differential. by opening and closing the bypass valve.
• During a test, we lowered  the differential pressure to 4psi. As it passed through 6psi  we got the evaporator loss of water flow error...this was the error that we had been having.. the problem was that the bypass value had been disconnected.
  

<- page up
 home_~phil