Welcome! Here are the website rules, as well as some tips for using this forum.
Need to contact us? Visit https://heatinghelp.com/contact-us/.
Click here to Find a Contractor in your area.

Calculate BTUs of Variable Speed Heating Loop

Hello all, after browsing this forum I believe I've come to the right place for help.
I have a large primary secondary heating system and with five secondary loops, each having a VFD controlled pump.The secondary loops are mostly 4" lines well over 1000' long. The secondary loops are essentially primary loops further downstream, so the flow rate is consistantly relevant to pump speed. A PLC monitors the secondary loops supply and return temps and controls the VFDs to maintain a specified delta tee. The PLC logs the power consumption of the pumps but now I'm looking for an economical way to get a general idea of the BTU-hours on each seconday loop. After doing mind numbing research I realized doing calculations with water is a lot more complicated than electrical calculations. The difficult part is getting the gpm, installing flow meters would run into a lot of $$$. Can pump head in a closed loop be calculated by (pump out psi - pump in psi) then take that number to the pump-curve chart to get gpm? Say I know my gpm at 100% pump speed. Is there a formula to factor in speed to get gpm at other speeds? I don't need a very accurate figure just a general idea. I'm not afraid of complicated equations I just need to get steered in the right direction. Sorry if this got long.

Comments

  • pecmsg
    pecmsg Member Posts: 4,775
    fans.....
    BTU's = CFM X Delta T X 1.08
  • Jamie Hall
    Jamie Hall Member Posts: 23,159
    Or gpm x delta T x 500...

    To @djelectric 's question. Yes and no. That is, if you know the pump curve at a specific rpm, and the pump head gain (psi out - psi in) then you can determine flow remarkably accurately -- assuming the pump is in decent shape. Unfortunately, the flow does not vary directly with rpm; the whole curve shifts. However, that said... some manufacturers publish curves with output at different rpms, and it is quite safe to interpolate between curves (it isn't entirely accurate, but close enough). Lacking that, with a little experimentation with a pump, one can measure the actual flow at some particular rpm and at several different head gains, and simply shift the full rpm curve to fit those points (I would take at least four points -- shutoff head, fully open flow (no head gain) and two others in between). Depending on the size of the pump, I've done this several times with two pressure gauges, a nice globe valve, some appropriate pipe, and a couple of buckiets (little pumps!) to 55 gallon drums (bigger pumps) to in one case a couple of 500 gallon tanks (pretty good size pump) and a stopwatch.

    Improvise, Adapt, and Overcome...
    Br. Jamie, osb
    Building superintendent/caretaker, 7200 sq. ft. historic house museum with dependencies in New England
    djelectricSuperTech
  • djelectric
    djelectric Member Posts: 5
    edited February 2019
    Thanks for the replies so far. @Jamie Hall I would have to put the loops out of service to remove the pumps for water barrel test, plus they're very heavy. After sleeping on it, I think my best bet is using the pump suction and discharge pressures, each pump already has the gauges. I also have the pump curve charts, so I think I can get a pretty reasonable figure. Now I'll create an example. there's a 15 psi variation=34.6 Feet Head. Say I have 1000' of 4" fiberglass pipe with 8 90deg elbows, and 180deg water. How do I calculate what my flow rate is?
  • hot_rod
    hot_rod Member Posts: 22,018
    An accurate gauge will be needed, ideally one gauge reading both sides to eliminate gauge error.
    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
    SuperTechGordy
  • Gordy
    Gordy Member Posts: 9,546

    Thanks for the replies so far. @Jamie Hall I would have to put the loops out of service to remove the pumps for water barrel test, plus they're very heavy. After sleeping on it, I think my best bet is using the pump suction and discharge pressures, each pump already has the gauges. I also have the pump curve charts, so I think I can get a pretty reasonable figure. Now I'll create an example. there's a 15 psi variation=34.6 Feet Head. Say I have 1000' of 4" fiberglass pipe with 8 90deg elbows, and 180deg water. How do I calculate what my flow rate is?

    Simply go to the pumps curve chart, and find 34.6 in the vertical head column go across until it intersects with the curve, and go down to the gpm horizontal line, and that’s the flow rate.

    The 15 psi differential pressure is the head pressure that pump is seeing at that flow rate of that loop so it’s irrelevant of the feet of pipe, and elbows etc. that’s all in the differential pressure reading.

    However one gauge as Bob noted is the best way to do it. So you eliminate gauge in accuracies.

  • djelectric
    djelectric Member Posts: 5
    edited February 2019
    Thanks @Gordy and @hot_rod , that should be easy enough. Assuming the pump curve is accurate, that should give me a good idea of the heat load. I can add all loop loads together and whatever deviation I have from the boiler output, I can put in an additional factor.. Just out of curiousity how would I calculate the pipe head loss? The way I understand with 180deg water, the Hazen-Williiams formula is inaccurate.


  • hot_rod
    hot_rod Member Posts: 22,018

    Thanks @Gordy and @hot_rod , that should be easy enough. Assuming the pump curve is accurate, that should give me a good idea of the heat load. I can add all loop loads together and whatever deviation I have from the boiler output, I can put in an additional factor.. Just out of curiousity how would I calculate the pipe head loss? The way I understand with 180deg water, the Hazen-Williiams formula is inaccurate.


    Depends on how accurate of a number you want. The tighter the number the more steps involved. If you can see and measure all the piping and fittings use the EL method.

    Other methods have you assume pipe length and use a factor to add for fittings, valves, etc.
    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream
  • djelectric
    djelectric Member Posts: 5


    hot_rod said:


    Depends on how accurate of a number you want. The tighter the number the more steps involved. If you can see and measure all the piping and fittings use the EL method.

    Other methods have you assume pipe length and use a factor to add for fittings, valves, etc.

    Thanks @hot_rod , I saw your screenshots come from idronics, and coming to there website I discovered loads of information . Thanks for all the replies in clarifying the GPM process to me.

    SuperTech
  • EBEBRATT-Ed
    EBEBRATT-Ed Member Posts: 15,452
    A lot of the new VFDs used on pumps are using sensorless technology. No differential pressure sensor. The pump curve gets programmed into the vfd the vfd senses the pump amps and can compute anything you want. Patterson Pumps uses Schneider Electric drives and & Armstrong uses Danfoss
  • Gordy
    Gordy Member Posts: 9,546
    hot_rod said:

    Thanks @Gordy and @hot_rod , that should be easy enough. Assuming the pump curve is accurate, that should give me a good idea of the heat load. I can add all loop loads together and whatever deviation I have from the boiler output, I can put in an additional factor.. Just out of curiousity how would I calculate the pipe head loss? The way I understand with 180deg water, the Hazen-Williiams formula is inaccurate.


    Depends on how accurate of a number you want. The tighter the number the more steps involved. If you can see and measure all the piping and fittings use the EL method.

    Other methods have you assume pipe length and use a factor to add for fittings, valves, etc.
    To me wouldn’t psi differential be more accurate other than gauge tolerance? It is what the pump is seeing for head.

    There is a lot of variables with EL calculations mostly pipe wall condition, and fitting connections.

  • djelectric
    djelectric Member Posts: 5

    A lot of the new VFDs used on pumps are using sensorless technology. No differential pressure sensor. The pump curve gets programmed into the vfd the vfd senses the pump amps and can compute anything you want. Patterson Pumps uses Schneider Electric drives and & Armstrong uses Danfoss

    these are the low end drives from Schneider Electric so no such features
  • hot_rod
    hot_rod Member Posts: 22,018
    Gordy said:

    hot_rod said:

    Thanks @Gordy and @hot_rod , that should be easy enough. Assuming the pump curve is accurate, that should give me a good idea of the heat load. I can add all loop loads together and whatever deviation I have from the boiler output, I can put in an additional factor.. Just out of curiousity how would I calculate the pipe head loss? The way I understand with 180deg water, the Hazen-Williiams formula is inaccurate.


    Depends on how accurate of a number you want. The tighter the number the more steps involved. If you can see and measure all the piping and fittings use the EL method.

    Other methods have you assume pipe length and use a factor to add for fittings, valves, etc.
    To me wouldn’t psi differential be more accurate other than gauge tolerance? It is what the pump is seeing for head.

    There is a lot of variables with EL calculations mostly pipe wall condition, and fitting connections.

    probably, I meant the more steps in the manual calculation, the more accurate them number, like the type of fluid, temperature of the fluid, actual fitting count, etc. Using this calculation method in lieu of having gauges.

    Ideally a certified, calibrated differential pressure gauge would give the most accurate pressure reading. Gauges like that can run upwards of a thousand bucks.
    Bob "hot rod" Rohr
    trainer for Caleffi NA
    Living the hydronic dream