I don't understand how (8) 600' 1" circuits, connected to a trunk line flowing 21.4 gpm, results in 37 fh.  Can you help?  If the 2 pumps are producing 40 fh, and the 8 600' circuits are consuming 37 fh, that only leaves 3 fh for everything else.  Yet we know the coax HE is using 22 fh because we measured 9.6 psi (* 2.31 = 22 fh).

Wish I was smart enough to figure this out.



When both units are running, as I wrote earlier, the measurements were:

5T: 63.1 In, 55.0 Out, 8.1 delta psi = 19.4 gpm through 2 pumps (from WF chart)
3T: 67.9 In, 63.5 Out, 4.3 delta psi = 10.8 pgm through 1 pump
Total flow out to borehole field = 19.4 + 10.8 = 30.2 gpm



Overall, replacing three UP26-116F pump motors with one UP26-99F (inserted into the 2" header line) would be very attrative to do.  Please help me understand how it would work - this is a pretty dramatic reduction in flow volume.

Where on the pump curve would the 3 scenarios operate (3T/5T units On, 5T On only, 3T On only)?

From your comments above, it looks like you're saying:

3T/5T On:

20 gpm = 2.5 gpm/ton
pump produces 16 fh (from UP26-99F curve) at 20 gpm
16 fh / 2.31 psi/fh = 6.9 psi = total of all friction loss in system, including 2 coax HEs in parallel


5T On / 3T Off:

? gpm
pump produces ? fh (from UP26-99F curve) at ? gpm
? fh / 2.31 psi/fh = ? psi = total of all friction loss in system, including 5T unit's coax HE


3T On / 5T Off:

? gpm
pump produces ? fh (from UP26-99F curve) at ? gpm
? fh / 2.31 psi/fh = ? psi = total of all friction loss in system, including 3T unit's coax HE




Many thanks!

Best regards,

Bill

The loopfield is about 1/3 of the pressure drop, the unit and the inside piping about 2/3. If you have only one pump, we usually have high CV zoning valves to close, so the variable speed pump revs down. However in your case you could just have run the water to run through both heat exchangers, even if only one pump runs, it does not matter much with a constant speed pump. 2 things are off in the calcs, it has 25 gpm going through the 1.5" pipe, and the header for the loopfield is quite long, and not accounted for. But the effect should be minimal, and you have room to spare.

Docjenser, awesome!  I sincerely appreciate your efforts.  After carefully studying the attachment, I now understand a whole lot more about how hydraulics works for closed loop systems.  Wow!


My conclusions:

1. WaterFurnace Envision 2-stage units need at least 2.25 gpm/ton; recommended flow is 3 gpm/ton.  Thus, for me, noting I have Envision ND064 and ND038 units, I need to move 25 gpm total - 16 through the ND064 and 9 through the ND038.

2. If my closed loop is modeled with the pump(s) in the supply/return line(s) (which means water is always flowing through both units regardless of which one is on/off), then I need a pump capable of 25 fh at 25 gpm.



Economic alternatives:

3. Looking at constant speed pumps, if I draw their various pump curves on the system curve you provided, it's easy to see that one UP26-116F (I have 3 of these at the moment) is marginal.  It will run at 22 gpm / 20 fh, which is 2.6 gpm/ton, a little bit above the 2.25 minimum noted above.  Cost to do this is moving one of my -116F pumps to the supply line, and close everything up.  Savings is $118 (my current setup) - 2419 hrs @ 335 watts @ 9.1¢ = $44 annual savings.

4. If instead I choose to put 2 of my three -116F pumps in the supply line, then I improve the system to run at a nice 3.1 gpm/ton, versus the 2.6 gpm/ton in 3.) above.  But, this will cost me.  Savings is $118 -  (2*2419 hrs) @ 335 W @ 9.1¢ = -$30 annual savings: i.e. a loss each year.

5. If I look at purchasing a -99F pump and trying to use just one, it gets to be very marginal.  Curve intersection is at 20 gpm at 16 fh, which is almost at the minimum.  Cost is purchasing a pump, opening up the supply line for the pump, removing the 3 pumps, and closing everything up.  Savings is $118 - 2419 hrs @ 245 W @ 9.1¢ = $64 annual savings.  And putting in two -99F pumps causes a loss again, just as in 4.) above.

6. A variable speed pump, one that can handle 25 gpm at 25 fh is going to be over $2000.  Still, it's operating cost, @ 100 W, is tiny, at $22/yr.  Annual savings is $118 - 22 = $96/yr, and payback years is probably around 30.  Maybe 3 year payback investment is aggressive, but approx 30 years is way too long.


Conclusions:

7. As a pump breaks, replace it with -99 models, which changes the running power from 335 to 245 W per pump.  Energy related savings is $11/pump/yr.  Making the change at replacement time means savings is immediate.

8. Attempt to model the system with the 5 ton unit having with just one -116F pump attached to it.  Figure out if total flow is at least 25 gpm when both units are running (it's 30 gpm right now).  If so, remove 1 of the 2 pumps servings the 5 ton unit.  Cost is connecting together the two flange fittings where the pump existed.  This may be total cost if loop does not need to be opened up and thus avoiding having to do a loop purge.  Savings is $33/yr.

9. Any other alternative has very long economic payback, espcially noting they require a service call with HDPE fusion tool and purge cart equipment familiarity.


Docjenser, much appreciate your efforts!  You've signifianctly advanced my knowledge of hydraulics theory, and how closed loop systems work.  And I have a clear idea of which alternatives are reasonable, and which ones are not.

Many thanks!

Best regards,

Bill

You really only need the minimum requirements. Your unit will run fine with in second stage. Keep in mind that your specific system is very much oversized given your short run times. So most of the time it is running in 1st stage only, further cutting down the flow requirements! One 26-99 would serve you well. You are simply riding the pump curve. But why are you looking at $2000 for a Wilo and why is that a 30 year payback?

Posted By docjenser on 24 Aug 2012 01:17 AM
You really only need the minimum requirements. Your unit will run fine with in second stage. Keep in mind that your specific system is very much oversized given your short run times. So most of the time it is running in 1st stage only, further cutting down the flow requirements! One 26-99 would serve you well. You are simply riding the pump curve. But why are you looking at $2000 for a Wilo and why is that a 30 year payback?

Docjenser, much appreciated!  You gave me new material to consider and learn.

The most heating or cooling BTUs that I need for this house is about 800 KBTU for a day at design outside temp (but at 77 degrees tstat setting, not Manual J 75 degrees), and that it will always be on the cooling side.  (See real time load).  It also shows that the heating side is not a consideration - a worse case heating day only needs about 350 KBTU (tstat at 68 degrees).  (I'm in a severe cooling dominated climate here in Dallas).

Also, I don't have any reason to believe that my Latent load needs depart from average conditions.  My inside RH (see chart) generally stays at 40 - 46%, with ocassional excursions up to 52% in Fall or Spring periods when little A/C is running and it's raining outside.

And lastly, my units never run in 2nd stage.  The only periods of 2nd stage are when the tstats are suddenly set to a setpoint so different from ambient such that the tstats' 20 minute 'upshift' timer kicks in.

So, I think there's plenty of quantitative evidence to conclude that I've got plenty of capacity with simply first stage alone.  And that's 2/3 * (3.2 + 5.3 tons) = 5.7 tons.  This I believe is what you're saying should be my "minimum requirements," not the 8.5 ton (total) rating of my 2 units.

And with a borehole field designed for 8 tons the field is more than adequately sized for 5.7 tons, and thus with respect to pumping closed loop water, the loop field is just not a factor.

So, relooking at the economic alternatives:

3.) Move 1 of my 3 UP26-116F pumps to the loop supply line, remove the other 2 UP26-116F pumps:  this will run at 22 gpm / 20 fh, which is now 3.9 gpm/ton, almost perfect.  Don't know what cost would be - no pumps to purchase but plenty of labor, including HDPE pipe fusion work, and a loop purge process needed.  Savings is $44 annually.

Can you give me a 'ball park' estimate that an HVAC professional would charge to do this: remove these 2 pumps, remove this pump, put in three 1" HDPE pipe splices about 6" long where pumps used to be, cut open 2" supply header, and install 1 of the pumps in the supply header.

4.) Move 2 of my three -116F pumps to the loop suppy header: no longer needs to be considered since 1 pump (3. above) meets recommended flow requirement.

5.) Same as 3. above but instead use a variable speed pump.  This is the alternative I remain most desirable of if economic numbers can be reasonable.  Savings is still $100/yr.  Labor cost is same too.  But an additional cost here is a new pump.

If my estimate of $2000 or more for the pump is too high, then can you give me a 'ball park' estimate for purchase of a Wily Stratos variable speed pump pump capable of 22 gpm / 20 fh?  Add to this the charge for the labor already outlined in 3. above, and dividing by $100/year, what is the payback period?

Many thanks!

Best regards,

Bill

You are rejecting 80 KBTU/H maximum, that is rejection, with the compressor heat, meaning that your house load is about 65 KBTU/h max, which is a 5 ton system.

Thus your are running in 1st stage most of the time, and your run time is pretty low, otherwise your circulation pumps operation costs would have a higher impact.
So the recommend flow is 3 gpm/ton, the minimum flow is 2.25 gpm/ton, which is 1.5 gpm/ton in first stage. For your system, if you run with 2.25 gpm, that is plenty and can be done by a single 26-99. Worst case,your unit runs in second stage and your delta T goes up be a couple degrees, not a big deal at all. Your EWT is not going to change.

This whole thing might not worth much to you economically, since you make your own power. I don't know. I can run the pressure drop more precise, I have Wilo Stratos seen to work well which were $500 (1.25x3-25) and might be good enough for your circulation requirements. May be a Grundfos ALpha is the way to go. I'd say you buy the pump for $500 and it should be done in 4 hours, $300 labor, purging, all together should be less than $1000 for a small wilo stratos, savings of $ 100, 10 year payback, 10% ROI. I would add two high CV zone valves, to slow the circulation pump down if only one heatpump is running. Extra material is $300.

Posted By docjenser on 25 Aug 2012 02:00 AM
You are rejecting 80 KBTU/H maximum, that is rejection, with the compressor heat, meaning that your house load is about 65 KBTU/h max, which is a 5 ton system.

Thus your are running in 1st stage most of the time, and your run time is pretty low, otherwise your circulation pumps operation costs would have a higher impact.
So the recommend flow is 3 gpm/ton, the minimum flow is 2.25 gpm/ton, which is 1.5 gpm/ton in first stage. For your system, if you run with 2.25 gpm, that is plenty and can be done by a single 26-99. Worst case,your unit runs in second stage and your delta T goes up be a couple degrees, not a big deal at all. Your EWT is not going to change.

This whole thing might not worth much to you economically, since you make your own power. I don't know. I can run the pressure drop more precise, I have Wilo Stratos seen to work well which were $500 (1.25x3-25) and might be good enough for your circulation requirements. May be a Grundfos ALpha is the way to go. I'd say you buy the pump for $500 and it should be done in 4 hours, $300 labor, purging, all together should be less than $1000 for a small wilo stratos, savings of $ 100, 10 year payback, 10% ROI. I would add two high CV zone valves, to slow the circulation pump down if only one heatpump is running. Extra material is $300.

I apologize for the delay getting back here.

Yes, the max Heat Rejection (HR) (structure and system) is 80 kBTU/hr - this can be seen quite clearly on this chart (peaks of red line).

Thank you for the reminder that this is not the same as Total Cooling Capacity (TC) (and in fact requires air flow measurement in order to accurately compute).  Yep, looking at the WaterFurnace Envision charts, indeed 65 kBTU/hr looks about right to estimate TC from 80 kBTU/hr HR.

So, yes, the structure needs about a 5 ton system, which is about exactly what I've got in first stage capacity (5 + 3 = 8 / .67 = 5.3 tons).

Noting 5 tons is exactly the capacity being applied to the structure, run times are still low because tstats are set to 77 day / 81 night (and visa versa for sleeping areas), versus the Man J and S default of 75°F.

Thank you also for pointing out that recommended flow at 3 gpm/ton is at 2nd stage, and thus obviously recommended flow is two-thirds (2 gpm/ton) in 1st stage (and indeed the minimum flow requirement goes all the way down to 1.5 gpm/ton).

So now I can see that yes, a single 26-99F is more than adequate for this system in Dallas temp conditions and the tstat settings in use.

WRT to cost of power: even if I didn't make any of my power, the last 12 month average for grid purchased power for me is 9.9¢, still a low enough number to not materially affect the economical analysis.  (I produce about one-third of total electrical energy need; for what I produce it cost me 6.9¢ across the last 12 months; thus my net 'mixed' cost across the last 12 months is the 9.1¢ I've been using for this economic analysis.)

So here's my revised summary, thanks kindly to your help:

1. When I'm ready to invest about $1300, the analysis done here over the past couple of weeks shows that I can change the pumping system to use a single Wilo Stratos variable speed pump motor inserted into the header pipe (before serving the two GSHP units), along with a couple of CV zone valves.  Pump operating cost is about $22/yr, way down from my current $118 annually, for an annual savings of $96/yr.  If we call this about $100, and we assume electric rates rise over the next decade, then payback period is probably in the neighborhood of about 10 years.

2. Meanwhile, the analysis done here clearly shows that if/when any of my existing three 26-116F pumps break, I've got so much pumping capacity that I can easily just pull out the broken pump and continue as is.

Really appreciated the assistance, docjenser!

Best regards,

Bill

10 year payback, that is 7% ROI for 10 years to go, guaranteed, argumentatively better than Wall Street.

10 year payback, that is 7% ROI for 10 years to go, guaranteed, argumentatively better than Wall Street.

10 years is sort of my make or break payback horizon. Under good- over bad. One of the reasons is any man made appliance is on the short side of life after 10 years.

Thanks!

Best regards,

Bill