Looking for help with some design questions for submersible pump

[chickenranch]

Hi, first time here but I've been reading the forums for a while and doing my own research.
I am designing and installing a system for a well shared between my property and a neighbor's. The big picture plan is to have a pump in the well that then feeds a cistern on each property, with a pressure boosting system coming from said cistern for the needs of each property.
Right now I am focusing on the system in the well itself.

The well information:
Total depth 320'
Static water 116'
Cased with steel 0-20 and 6" SDR17 PVC 20-280'
Open hole 280-320'
Pump depth recommended by driller: 200'
Don't know ph but the water is hard.

The aquifer is large and stable. The drillers didn't do a proper pump test, instead they air lifted it for 9 hours and recorded 50+gpm. It is my understanding this method provides no reliable estimate of sustained yield. A neighboring well was tested at 7gpm and had 5ft drawdown over 1.5hr.
The rock is Navajo sandstone, and the drillers said they normally case to 200' and then have open hole after that, but in our case the sandstone was too soft all the way down to 280'. They went down a little further and hit a big soft pocket, which they blew out for hours, producing lots of muddy water until it was fairly clear. Then they went a little deeper and it was clear but producing some sand, which doesn't seem surprising given all the above. This was at the end of April 2025.

Given this, it has been my design goal to pump at a conservative rate of 10-15gpm in order to be gentle on the formation, using the cisterns to buffer high use flow for garden irrigation. I also like that the cisterns mean the submersible pump will run at most twice a day, and will probably only turn on a few times a week during periods when water is only being used in the home.

The piping system goes something like this:
180-200' up from the pump, splits at a tee and (1) runs 330' to cistern 1 with 6' elevation gain and (2) 250-350ft to cistern 2 with ~20' elevation gain. The cisterns are not sealed. I have the TDH at 12-15gpm somewhere in the 150-170 range depending on how you calculate fitting friction head additions using 1.25" SIDR HDPE pipe.

As for the pump, I have been planning on going with either the Grundfos 10SQ-07-200 (for a 11-12pm flow) or the 15SQ-07-150 (15-16gpm flow). I like the soft starting feature, but it may not matter given how infrequently the pump starts - that said, the electrical cable will be about 500' in all so keeping the starting surge 1x rather than 6-10x might be good for voltage stability. It is also my understanding that a soft start can be gentler on the formation if it is slightly weak/soft.
I'm also interested in the 23000 or 24000 series at AY MacDonald, very curious to hear people's thoughts on these pumps vs grundfos.
Planning on 1" 200PSI drop pipe.
From reading the forum it sounds like I should install a flow sleeve. The well water is in the 50s, but the pump will also be running for 4-7 hours at a time.

So, here is my list of questions, and in general I am interested in any thoughts people have on this setup and things I am missing.
1. For this well, where would you set the flow rate to best protect the formation?
2. What is a realistic timeline I should be designing around for the lifetime of parts in the well? 15, 20, 30 years? I am down to spend a little more up front for a system I don't have to touch again until I'm 60.
3. What material should I use for the fittings? I am getting conflicted information on SS 316 vs 304 vs brass for the pump to pipe connection in the well. What will handle corrosion best in hard water with (maybe) moderate chlorides for 20-30 years?
4. Should I use a support line? I know people here with experience feel like safety lines inevitably fail and end up making it extremely difficult to pull the pump, but I wonder if poly alone can safely support the pump assembly in this setup. I think the weight of cable, pipe, and pump come to something like 70lbs at the most. I was advised to use a 316 SS wire cable - it's sort of hard to imagine this coming undone with thimbles and swages, but maybe it could?
5. I was reading about a slow commissioning process for the well to help harden/stabilize the formation to prevent it from collapsing over time or producing lots of sand. Curious if anyone has experience or insight to offer on this process.

Thanks for your time!

 

[Valveman]

Controlling the well pump to supply two cisterns will be a concern. I would use electric solenoid valves at each cistern and use a Cycle Stop Valve and small tank/pressure switch to control the well as needed. Cable, rope, bad idea. 200# poly is tuff stuff. Just make sure to use SS of Brass long barb fittings with at least two hose clamps each, buckles opposite, and taped over with electric tape to prevent the screws from rusting out. 400 series SS hose clamp screws will rust. SQ is a good pump but so is the McDonald. A 10 GPM well pump is more than needed as the cisterns provide lots of cushion.

 

[Sponsor]

 

[chickenranch]

Yeah I've been working through the two cisterns 1 well issue and it basically seems like I will need float switches, solenoids, and a contactor/relay with some logic to have it work. So it goes, cheaper than paying for the well myself. It's 100/ft here.

On the cable/rope piece, I guess it's not really the poly that is in question here but the ultimate integrity of the clamps. I was going to use 316 SS for these since that's not hard to find and then use the extra long barbs with 3 or maybe 4 haha.

So you don't think SS or brass really matters as far as relative corrosion risk. It has been recommended to use 316 since it handles corrosion and pitting better than 304, but also that brass is even better for corrosion in a mineralized, low oxygen environment. But is galvanic corrosion a bigger threat? I realize I'm pretty deep in the weeds here, but looking for thoughts based on experience and observation rather than theory.

As far as the flow rate, it really depends on the size of the cistern and worst case use patterns. I haven't gone super deep into gaming that out yet.

Don't see the need for a csv before the cisterns, seems like wasted money. Why bother? I don't care about the pressure here.

EDIT: Okay I didn't look very carefully at the diagram, but I see you have an outlet flow to a house before the cistern so in that setup, the csv makes sense. I could consider this but I don't think I want the submersible pump being called on to meet home loads directly, since it increases the number of starts dramatically and also imposes another 140ft of head on the submersible pump that means I need to size it up.
And I would probably need to put that setup on the house end of the distribution pipe since there is no plan to have a pump house at the well head, which would mean that if the other cistern calls for water at the same time House 1 is using water, House 1 suddenly has a trickle which is no good haha.

 

[Valveman]

You don't need "logic" or even wires running from the cisterns to the well pump. Just put an electric solenoid valve on each cistern that opens and closes as needed to fill the cistern. The CSV on the well pump allows the use of a small pressure tank with a pressure switch, and the CSV will vary the flow to either one or two cisterns as needed. The solenoid valves at the cisterns should have flow control knobs so you can adjust the flow rate filling each cistern. With a 10 GPM well pump each cistern should be limited to fill at 5 GPM. Without the CSV the well pump would cycle on and off when filling only one cistern at a time. There are many other benefits to the well and pump when using a Cycle Stop Valve. Why bother? Maybe because of 50+ years experience and observations. Real brass fittings are good. But SS is better than the low lead version of brass.

 

[SuperGreg]

My well system pumps into a pressure tank, through a series of filters, then fills my cistern through a float valve. This way I trickle clean filtered water into the cistern. It would be trivial to tee off the output into another cistern with a second float valve. The requirement is the main pump needs a pressure tank and pressure switch, other than that it's pretty simple.

Just re-read Valveman's post and I think this is basically saying what he said. I chose not to use a CSV because I have a chlorine metering pump. If I could switch that to proportional based on flow then I would go CSV.

 

[chickenranch]

Okay I see the logic here, the pressure tank/pressure switch is basically just a way of controlling the well pump using pressure as a signal. That's a great idea.

In my situation, the well head is about 300' from the other infrastructure. I wouldn't be running extra wires to the wellhead, rather I would have the contactor by the power source and run signal wires from the cistern float switches to there. This would be easy enough if it was just my cistern, which is nearby the power, but the other cistern is hundreds of feet away and this would be really annoying to do.

So I actually love this idea of using pressure as the signal since the pipes are already there. Thanks for that insight.

The CSV seems unnecessary at the wellhead, however, simply adding artificial head and another part that can break/needs to be bought. Whether there is one cistern at the wellhead, or two cisterns on each property, I don't see any conditions that would cause short cycling. To my understanding, a cistern calls for water, the pressure drops and cut-in point is reached, the pump turns on, and runs at duty point on the pump curve continuously. Cistern is full, solenoid closes, pressure rises to cut-out point, and pump turns off. If the other cistern called for water sometime during this cycle, the only difference is the flow would split according to the head ratio between the two cisterns. So in this use case I don't see how adding a CSV adds any function, especially around short cycling. Am I wrong about the hydraulics here?

I do see that you both mention proportional flow, and I don't understand the role the CSV would play in enforcing this, but maybe that's where it's function comes in. How does that work, and why wouldn't the addition of two 5gpm flow reducers on the cisterns accomplish the same thing?

In my case, my math shows that the split between the two cistern would be something like 5:7 for 12gpm flow, which is okay with me, and no flow reducers also allows each cistern to fill faster if the other isn't also calling which seems like a nice benefit. But once again, open to reasons why this doesn't make sense or is a bad idea.

All that said, I am planning on using a CSV after the cistern for the pressurized side of the system. Fits the bill quite nicely on that end!

 

[Fitter30]

Would use two floats per tank. Top one for off and bottom for on. That way the on off is separated. Would also use motorised ball valves they don't slam closed like a solenoid. 300' 12 gauge wire @ 240 vac 5 amp load is eight under 3% voltage drop.

 

[SuperGreg]

Would use two floats per tank. Top one for off and bottom for on. That way the on off is separated. Would also use motorised ball valves they don't slam closed like a solenoid. 300' 12 gauge wire @ 240 vac 5 amp load is eight under 3% voltage drop.

Can you explain exactly how this would work? I'm familiar with fill/stop sensors but usually paired with a logic board.

 

[chickenranch]

Would use two floats per tank. Top one for off and bottom for on. That way the on off is separated. Would also use motorised ball valves they don't slam closed like a solenoid. 300' 12 gauge wire @ 240 vac 5 amp load is eight under 3% voltage drop.

The two float valves was recommended to me for the cistern, especially since the high and low will be feet apart from each other.
Motorized ball valve is a great idea, thanks for that.
I was going to go with 10awg because the run from power source to motor is 500' including the 200' down to the pump itself. Pump is rated at 8.4A which with #12 would be 6.3%, #10 is 3.8%. Grundfos can operate with up to 9% drop but I figure the extra $80 in cable is worth the pump being a bit more efficient and operating closer to its desired voltage.


Curious for thoughts from anyone on the hydraulic logic of the CSV at well head before the cistern compared to not having it there.

 

[Bannerman]

thoughts from anyone on the hydraulic logic of the CSV at well head before the cistern

Unless 100% of the pump's flow capacity is always consumed, then any restriction in flow will cause the pump's excess unused flow capacity to enter the pressure tank. Once the tank becomes filled to increase the system pressure to the pressure switch cut-off pressure, then the pump will be shut down. Once the pressure is again reduced to the pressure switch cut-in pressure, the pump will be restarted. This cycling will continue to be repeated for the entire time the consumed flow rate remains less than the pump can supply.

As a simplified example, you mentioned a flow rate of 12 GPM which I assume to be the pump's flow rate. If the flow rate entering each cistern is 6 GPM, and if both are being filled at the same time, then 100% of the pump's 12 GPM delivered flow will be consumed between both, and so the pump will run continuously without cycling. If however, only one cistern is being filled, then only 6 GPM will be consumed, so the excess 6 GPM will enter the pressure tank, and cycling will occur.

A CSV will reduce the pump's delivered flow rate to match the actual rate of consumption. If only one cistern is filling at 6 GPM, then a 12 GPM pump will only supply 6 GPM, thereby preventing the pressure tank from filling which will prevent pressure from further rising which will prevent cycling for the entire duration of water consumption. If water consumption increases or lowers, then the CSV will increase or reduce the flow rate to match that consumption rate, while continuing to prevent the pump from cycling. If water consumption increases to equal 100% of the pump's capacity, then the CSV will be fully open and will not restrict flow from the pump whatsoever.

Once no water is further consumed, the CSV will close to prevent further pressure increase downstream, but because the CSV is designed to permit 1 GPM to continue to 'leak past', water from the pump will continue to enter the pressure tank at 1 GPM until pressure increases to the pressure switch cut-off pressure, whereby the pump will then be shut-down.

A pump will consume the greatest amount of electrical power when moving water at full capacity. Because a CSV will reduce the flow rate from the pump when the flow rate is lower, the pump will not then need to move the maximum flow of water and so the power the pump will consume will typically be reduced, which will reduce the amount of heat the pump will produce.

 

[Fitter30]

500' 10 gauge wire isn't large enough for powering the pump only control voltage. Heres a wire size calculator. 3% max voltage drop. Diagram for floats have draw one up.
www.inchcalculator.com/voltage-drop-calculator/

 

[Reach4]

10 AWG wire is good for 760 ft each way for a 3/4 HP motor according to the Franklin AIM manual. So that would be the sum of the horizontal and vertical run of cable. With 1 HP, that would be 630 ft.

The Grundfos 10SQ-07-200 and the 15SQ-07-150 are 3/4 HP.

I do recommend a flow inducer sleeve. For suggestions on 3-inch pvc pipe to make the flow inducer, let us know.

 

[Fitter30]

The two float valves was recommended to me for the cistern, especially since the high and low will be feet apart from each other.
Motorized ball valve is a great idea, thanks for that.
I was going to go with 10awg because the run from power source to motor is 500' including the 200' down to the pump itself. Pump is rated at 8.4A which with #12 would be 6.3%, #10 is 3.8%. Grundfos can operate with up to 9% drop but I figure the extra $80 in cable is worth the pump being a bit more efficient and operating closer to its desired voltage.


Curious for thoughts from anyone on the hydraulic logic of the CSV at well head before the cistern compared to not having it there.

Will need 3- 10 gauge wires w/ ground minimum. Din rail relays 240 vac coils have 2 amps
Bottom float float energizes R2 pumps on. R3-1 holds pump on
Runs till R1 float makes tank full. R1-1 opens pump off
Level dropsR1-1 makes R2-1 makes turns pump back

on

 

[Valveman]

As was said, using an electric valve at the cistern and a pressure tank/pressure switch at the well will let you fill as many cisterns as needed without any wires, relays, etc.

A regular sprinkler or solenoid valve is very dependable and inexpensive. They don't cause water hammer if you get the kind with the flow control knob and adjust it to low flow as needed. Without the flow control knob an electric solenoid valve will throw itself wide open when filling a cistern and water hammer from closing fast. With a flow control knob, 6 GPM is barely open and there is no water hammer on valve closure.

A Cycle Stop Valve on the well pump will keep it from cycling itself to death when using low flow to fill a cistern. Without a Cycle Stop Valve the well pump will cycle unless you are filling enough cisterns with enough flow to use all the water the well pump can supply.

 

[chickenranch]

Wow thanks for all the comments, think I replied to everything below and there is also a question at the end about where the pressure tank/pressure switch can be located in the system IF there is no CSV. Once again would love the perspective of the experienced on this.

Just to be clear, I am totally on board for the pressure signal control system, I think it's a great idea and I am going to do it, but slightly differently than proposed because of my particular system.

Unless 100% of the pump's flow capacity is always consumed, then any restriction in flow will cause the pump's excess unused flow capacity to enter the pressure tank. Once the tank becomes filled to increase the system pressure to the pressure switch cut-off pressure, then the pump will be shut down. Once the pressure is again reduced to the pressure switch cut-in pressure, the pump will be restarted. This cycling will continue to be repeated for the entire time the consumed flow rate remains less than the pump can supply.

Okay I see where the issue is here. IF you decide to restrict the flow rate to each cistern to a point below the flow rate at a given TDH on the pump curve, you are in cycling territory and the CSV would stop that. Makes sense.

But if you don't restrict the flow, then the pump runs continuously, which is what I am planning to do because I would rather each cistern get as much as it can when it can, and I am okay with slightly different fill rates when both have their valves open. I am doing this in part to reduce the difference between flow into and flow out of the cistern during high use irrigation periods since this allows me to build a smaller cistern to meet my needs.

So we're basically in agreement, just working from different assumptions about the behavior of the system. Thanks for the explanation, that clarifies the concept for me.

[...]

A regular sprinkler or solenoid valve is very dependable and inexpensive. They don't cause water hammer if you get the kind with the flow control knob and adjust it to low flow as needed. Without the flow control knob an electric solenoid valve will throw itself wide open when filling a cistern and water hammer from closing fast. With a flow control knob, 6 GPM is barely open and there is no water hammer on valve closure.

A Cycle Stop Valve on the well pump will keep it from cycling itself to death when using low flow to fill a cistern. Without a Cycle Stop Valve the well pump will cycle unless you are filling enough cisterns with enough flow to use all the water the well pump can supply.

As far as water hammer goes, I had been thinking that the pressure tank might buffer some of that. Assuming a 12gpm flow rate and either a 1gal or 2 gal tank capacity, when that valve slams shut the tank would then fill in 5s or 10s, respectively (technically a little longer than that since we would move up the pump curve as pressure increased, but that is conservative). Is that not long enough to prevent water hammer? I have no idea but thought that would be okay.
And since the pump would be running for hours before this point, I'm not concerned about minimum run time.

If that 5-10s of tank filling isn't enough time to prevent water hammer, it looks like a motorized ball valve is the same price as a comparable electric solenoid, so that seems like it would deliver all the benefits without one more part (the CSV), and thus be less expensive. Not to mention that I can keep my less powerful, less expensive submersible pump rather than sizing up to handle the artificial head the CSV would create.

And @Valveman, since I have you here, I have a question I have been trying to figure out about the CSV125 and CSV1A that I can't find an answer to on the forums at your site. I was looking at the specs for these two models and I was wondering - first, why is the friction loss in the CSV125 so much less than the CSV1A? It is just the difference between the 1.25" and 1"? Second, are the pressure falloff and friction loss additive for the CSV1A, or should I refer to one chart or the other for figuring out where to set the CSV1A so I get my desired pressure at a higher flow rate?

Energy Use

[...]

A pump will consume the greatest amount of electrical power when moving water at full capacity. Because a CSV will reduce the flow rate from the pump when the flow rate is lower, the pump will not then need to move the maximum flow of water and so the power the pump will consume will typically be reduced, which will reduce the amount of heat the pump will produce.

While it is true that it is using more energy, it is also moving more water and the watt-hours per gallon is much better. What I have noticed in my research on both VFDs and CSVs for constant pressure systems, low flow is just a killer when it comes to efficiency for most pumps (like in the 1-3gpm range). I have been looking at pumps that can do 15-20gpm so that's probably part of the reason why I am seeing that.

I feel like this energy example is sort of based on a varying-flow-with-constant-pressure situation but not really relevant to the cistern fill situation, since with unrestricted flow you're chilling at constant pressure anyway. Like you said, the CSV is there to prevent cycling when you throttle flow into the cisterns. And I think it should be said that CSVs and VFDs don't save energy - they prevent short cycling, and you pay an energy cost to do that. How much you pay depends on how often you operate off the BEP, and how far off it you are. I think it's worth it, but it bears keeping in mind.

To illustrate:
Here is the pump curve for the 10SQ-07-200. You can see that at 6gpm, it is operating at 40% pump efficiency and using 1HP/745W. If you wanted to convert that to watt-hours per gallon, you end up with ~2.07Wh/gal.

Compare that to the duty point at 12gpm. We are now running at ~47% efficiency, pumping 12gpm using 1.12HP/834W. So twice the water production for only 12% increase in energy consumption. This comes out to 1.15Wh/gallon. T

So I could change my pump in, add a CSV, add flow reducing solenoids to the cisterns, and end up at a different place on a different pump curve and probably be at a similar BEP and Wh/gal. But then all I've done is buy a more expensive pump, an extra piece of hardware, and reduced the maximum fill rate of my cistern - if the goal is to reduce water hammer, it seems like a motorized ball valve would be a simpler fix. If the goal is to equalize the fill rate when both cistern valves are open, I could probably add some friction head to my cistern and get them close. Just seems like there are simpler and less expensive options to meet both of these goals.

That said, I can totally see how in a larger system with more cisterns, or really unequal partitioning of water based on TDH to each cistern, you would want to impose a controlled flow rate on each cistern, in which case the system you propose would be the most elegant and cost effective approach. I appreciate you sharing it with me, it has been good food for thought and has given me a simpler way to control my system.

My Question

I have been thinking about the pressure tank/pressure switch setup at the wellhead. I was making plans to build a little sunken box out in the field to access the PT/PS, ball valves to isolate the lines, and an small electrical panel with the contactor and a SPD (we have a lot of lightning, one more reason I am glad to be using pressure control and not logic relay).

Then I thought it would be nice to NOT build something extra way out in the field. What if I could instead have the line from the well split at the wellhead, and when it comes my way, route it into the pumphouse next to the cistern and have the PT/PS there, instead? Then it just controls the pump at the panel in the pumphouse, I am putting it in a structure I have to build anyway, and I can monitor it way more easily.

Thinking about the pressure signal, it seems like it would still work. Let's say both cistern valves are closed. My cistern opens, pressure drops, pump turns on. Other cistern opens, pump is still running, both fill. My cistern closes, other is still open, pressure stays below cut off point, pump keeps running and other cistern keeps filling. Other cistern closes, pressure rises to cut off, pump turns off.
Or, my cistern is closed, other cistern is open. Pressure drops (I guess it has to travel farther, would that make a difference?), pump turns on, other cistern fills. Other cistern closes, pressure rises, pump turns off.
So to my thinking it all still works, but once again this is theory to me, so what am I missing, if anything? Does the proximity to the valve at my cistern send weird pressure waves to the pressure switch that could cause chatter? Does the distance from the other cistern create problems?
And I know this won't work with a CSV, if I did use one I would put this all at the wellhead and get a bigger pump.

Thanks everyone, this has been extremely informative. I really appreciate the help thinking through all of this.

 

[chickenranch]

Will need 3- 10 gauge wires w/ ground minimum. Din rail relays 240 vac coils have 2 amps
Bottom float float energizes R2 pumps on. R3-1 holds pump on
Runs till R1 float makes tank full. R1-1 opens pump off
Level dropsR1-1 makes R2-1 makes turns pump backView attachment 107088 on

Thanks for this, I will bookmark this for later because I am not on this level yet! Also I love your handwriting.

 

[Fitter30]

Thanks for this, I will bookmark this for later because I am not on this level yet! Also I love your handwriting.

Valves in valve box that gets full of water no big deal. Electric not so much. Put the electric above ground its hard enough keeping. Nema 4 enclosure.

 

[Valveman]

I commend you on your research. I have been saying for 40 years that VFD's do not save energy and a CSV saves just as much energy as a VFD. Lol! You are correct that any pump is most efficient at its Best Efficiency Point or BEP. However, there are still many advantages to using a CSV. You still need to use the pump as close to BEP when possible, especially for high use applications. But when the system needs less than the pumps max flow, which is most of the time, a CSV is very beneficial.

You also do not need a larger pump to work with a CSV. That pump makes much more flow and pressure than needed, as pumps need to do for when max flow is needed, and the CSV will make it work at max flow when needed and down to 1 GPM as needed as well. That pump needs at least 25 PSI restriction, like a Dole valve, to keep from pumping more than 12 GPM from 116' deep. I would think anything close to 15 GPM would cause upthrust problems. Pumps are made to make pressure, so back pressure is good for them. Just restricting the pump to 6 GPM where it will draw 745 watts will make it run cooler than when pumping 12 GPM at 834 watts. It maybe less efficient, but will last longer, and that needs to be figured into efficiency as well.

Ok now to the problem you haven't seen yet. With both cisterns open to 12 GPM, the system pressure will fall to zero when both are filling at the same time. Unless you use a Dole valve or something to restrict the pump, it will be splitting 15+ GPM into both cisterns, and the pump will be in upthrust. Not only that but when you go below the air charge pressure of a diaphragm tank of any size it makes water hammer. Then when the cisterns are full and the both valves close, the system will water hammer again as the pressure hits the air charge in the tank. This water hammer is usually strong enough to bounce the pressure switch on/off rapidly. ESPECIALLY with a small tank. Did I mention the damage to the check valve that happens every time it slams shut form 12-15 GPM?

Yes, a pump is most efficient at the lowest pressure. But pressure is good for the pump and makes everything work properly. Lifting from 116' that pump will make 75 PSI pressure at the surface. That is plenty to use a 20/40 pressure switch. Adding a CSV to hold 30 PSI will let you use a small pressure tank and still get the recommended run time. With each cistern set to accept 6 GPM, the CSV will put 64 PSI back pressure on the pump when filling only one cistern and 24 PSI back pressure when both cisterns are filling. This will keep the system above 20 PSI, so it doesn't hit the 18 PSI air charge in the tank and water hammer. The CSV will also put 75 PSI back pressure on the pump as it is filling the pressure tank. This causes the tank to fill at 1 GPM, the motor to cool down at really low amps/watts, and the check valve to close gently when the pump does shut off.

As to which CSV to use, CSV1A or CSV125, it will make little difference in this case. The CSV1A will be best because it is adjustable and you need it set for 30 PSI. The only time it will be at less than 6 GPM is for the minute it is filling the pressure tank before shut off. The CSV will work no matter where you put the pressure tank, but the CSV needs to be on the well line before it tees off to the other cistern.

The Irritrol solenoid valves I use with the flow control knob are only like 30 bucks. I don't know how you flow control an electric ball valve, but I guess it could be done electrically.

While the CSV will be helpful with the well pump, it is just as helpful with the booster pumps, and they will deliver that strong constant pressure to the houses.

Hope I didn't miss anything. But there are always unseen uses and circumstances with a water system. The whole point of the CSV is that you don' have to get everything perfect. The CSV will take care of the pump and deliver strong constant pressure and let you do just about whatever you want with a water system. A CSV system is fairly fool proof. Once it is set up you can use water from just one faucet or all the faucets at the same time without worrying what it is doing to the pump.

 

[Bannerman]

But if you don't restrict the flow, then the pump runs continuously, which is what I am planning to do because I would rather each cistern get as much as it can when it can, and I am okay with slightly different fill rates when both have their valves open. I am doing this in part to reduce the difference between flow into and flow out of the cistern during high use irrigation periods since this allows me to build a smaller cistern to meet my needs.

To prevent the well pump from cycling even while supplying flow to only one cistern, without a CSV, the unrestricted flow capacity to each cistern, will need to equal or exceed the maximum flow rate the pump is capable of delivering.

Since water will always take the path of least resistance, this creates a potential issue when both cisterns require refilling at the same time. The cistern located the shortest distance from the well pump, or the cistern placed at a lower elevation, may receive 100% of the well pump's flow whereas the higher elevation/more remote cistern could receive 0 flow. A temporary lack of supply to one cistern, may not create much of an issue under most circumstances, but if both cisterns are supplying a large volume of water for irrigation at the same time, then one cistern may become fully depleted, even as the well pump continues to fully supply water to the alternate cistern.

 

[chickenranch]

[...]

You also do not need a larger pump to work with a CSV. That pump makes much more flow and pressure than needed, as pumps need to do for when max flow is needed, and the CSV will make it work at max flow when needed and down to 1 GPM as needed as well. That pump needs at least 25 PSI restriction, like a Dole valve, to keep from pumping more than 12 GPM from 116' deep. I would think anything close to 15 GPM would cause upthrust problems. Pumps are made to make pressure, so back pressure is good for them. Just restricting the pump to 6 GPM where it will draw 745 watts will make it run cooler than when pumping 12 GPM at 834 watts. It maybe less efficient, but will last longer, and that needs to be figured into efficiency as well.

Okay this is a good thing to check. Here's my math on TDH for the pump curve. Pump is at 200', dynamic water level around 120'. Cistern 1 will be about 250' from the wellhead and 4' higher. (Cistern 2 location is uncertain but 250-300' away and ~20' higher).
So I have 120' + 4' + 450' of 1.04" HDPE pipe run. Obviously friction changes with flow rate so I just have to check back and forth on that, and putting it through a friction calculator using Hazen-Williams 12.5gpm gets me about 39' of head, so I'm at 163' total. This lands me at more or less 12.5gpm on the 10SQ-07-200 pump curve.

But this doesn't actually account for the friction losses in the fittings. I feel like when I work with this I get a range of values and would love a source you think is good for calculating this. Without the CSV and the pressure tank/switch in a valve box deal at the wellhead, my fittings look something like this for each cistern: long barb, long barb, pitless adapter, barb, ball valve, T that water flows through for PT, ball valve, T that splits the water, ell, ell, pass through for the cistern.

I am getting values for equivalent length all the way from 50' to 126'. If I take the low end then I am working with 124' + 500' equivalent pipe length, at the high end 124' + 576' pipe length. Messing around with the friction calculator and the pump curve, that puts me at 12.4 gpm @ 166' TDH at the low end (so basically no change), and 12gpm @170' TDH. This is for Cistern 1.
For Cistern 2 I could assume it's all the same, maybe plus a little more pipe run, plus 20', which would put me at 11.5gpm @186' TDH. These numbers are for a single cistern running at a time.

It's true that the pump CAN handle some extra back pressure - in these scenarios I am operating a bit to the right of BEP, and in the Cistern 1 scenario you could add 40' of head / 18 psi and be hitting 10gpm. And in fact, before either cistern is built I am planning to put a CSV on the well pump and just run it as if that was the whole system. I figured that I can add 40-50psi and still get 6-7gpm which will be okay for the time being. I suppose with the CSV1-A I could set it that low and hit the BEP.

I certainly take your point that energy efficiency isn't only about energy used, since if that was the only metric people should just use a normal pressure tank with the pump sized to be running at BEP the midpoint in the pressure differential switch. Doing some calcs it looks like you only lose about 20W more in the 12gpm case than the 6gpm case as heat in the motor and pump. Dunno if that really justifies the 2x energy use but what do I know. And at these flows with a flow sleeve and cool well water seems like it wouldn't matter a whole lot but *shrug*

[...]

Since water will always take the path of least resistance, this creates a potential issue when both cisterns require refilling at the same time. The cistern located the shortest distance from the well pump, or the cistern placed at a lower elevation, may receive 100% of the well pump's flow whereas the higher elevation/more remote cistern could receive 0 flow. A temporary lack of supply to one cistern, may not create much of an issue under most circumstances, but if both cisterns are supplying a large volume of water for irrigation at the same time, then one cistern may become fully depleted, even as the well pump continues to fully supply water to the alternate cistern.

Yeah I have no real world experience with this, but I'm trying to think about what would happen with a hose if you had a Y and had one section on the ground and the other section raised a bit. I guess I should just go try this and see what happens since it's basically just a lower pressure version of this situation. My intuition tells me I would still get flow from the raised hose, but less, and always something? What I am reading about this from the theory point of view is that the head loss need to reach equilibrium. If the pipes are the same length, same diameter, same fittings and cistern 2 is 16' higher, it looks like the flow would split 8/4 favoring cistern 1. This might be okay for this system with these users, but it's definitely a bit lopsided. Looks like I could equalize flow by sizing up to 1.25" pipe to cistern 2 and placing cistern 2 such that elevation difference was 9' rather than 16'.

But to Valveman's point, the benefit of the CSV is however this played out you would avoid pump cycling since it just handles whatever is happening. Definitely worth considering!