Increased back pressure and pump efficiency

[97044]

I am looking at using a cycle stop valve in an irrigation well system.   It is a deep well (well depth 700', guessing the pump is around 650', static water level is 250' above the pump or ~400' below ground) with an unknown 5hp pump down the well and will be used for irrigating a few different properties. 

Ideally we would match the pump flow to the size of the irrigation zones and have the irrigation controller switch on the pump.  However as we will have several different irrigation controllers attached (and want the use of spigots as well) this doesn't work.   So now I am considering a pressure regulator + bypass or a cycle stop valve.

The part I am having difficulty with is the efficiency of the system, particularly if we use lower flow drip irrigation zones.    The pressure up stream of the valve will always be what ever the ultimate pressure of the pump is (at the bypass flow).   Doesn't this force the pump run at the left most edge of its performance curve, which is way outside the optimum efficiency of the pump?  Does this cause issues with pump cooling?

[Cary Austin]

I am guessing you have a 25 GPM, 5HP because of the depth.  That pump can build a max head of 800'.  But because your static level is at 400', that only leaves 400' of head or 173 PSI at the surface.  A CSV3A2T needed for that size pump has a minimum flow of 5 GPM built into the valve.  5 GPM is more than adequate to cool a 5 HP pump/motor.  One of the reasons 5 GPM can keep the motor plenty cool, is because the amps will drop when the flow is restricted with the CSV.  Depending on the brand of pump, the amps will drop from 20% to 50% from full load.  This is the same amount of amp drop you would get using a VFD on that pump.  However, a VFD also waste energy at low flow rates.  Even though the energy use drops by 50%, the flow rate is reduced by 90%, and it is using more Kw per gallon produced.  ALL VFD's DO THIS AS WELL!  So, the CSV saves just as much energy as a VFD.

All pumps are most efficient when running at there best efficiency point BEP, which is almost at max flow.  Reducing the speed with a VFD or restricting the flow with a CSV will both use more energy per gallon produced.  Therefore, it is best to design every zone as close to the pumps BEP as possible.  Tie multiple drip zones together or run them at the same time.  The CSV will let you safely run any size zone you want above 5 GPM, but the closer to BEP the more efficient it is.

The CSV will allow the use of multiple irrigation controllers and hose bibs at any location.  The CSV also gives a mechanical soft start and soft stop when the pump does start or shut off.  The convenience of that and being able to run any amount of water, anywhere, anytime, while making the pump last several times longer than normal makes the CSV the best choice for controlling a pump.

[97044]

I was not considering a VFD as an alternative option, but a large pressure tank.   

In the pressure tank case the pump would be running against a back pressure 173 + 50 psi= 223psi as opposed to 173 + 173 = 346psi with the CSV.    The physics here is the part that is bothering me.   Regardless of the efficiency of the pump, the work the pump is doing is equal to flow * delta_Pressure.  So in the case with the valve it is doing 5x346 = 1730 units of work, whereas operating against a pressure tank it is doing (on average) 5x223 = 1115 units of work (the units don't make sense here, but they are the same).    There is really no way around the fact that a pressure reducing valve is wasting energy, it is reducing the energy of the fluid, without doing useful work, that energy has to come from the pump.

Also with the pressure tank option, the pump will cycle at a rate depending on the draw down of the pressure tank, but when it is on, it will be running close to the BEP, so it will be more efficient.    I am trying to balance increased cycling vs reduced efficiency.   Is running at ~65% efficiency worth the decrease in the cycling?

[Cary Austin]

You are on the right track.  But looking for the most efficient way to pump water I can't believe someone hasn't tried to convince you that just adding a VFD will save 50% in energy.  Lol!  I don't know how many times I have heard it.  But you are exactly right that there is no more efficient way to pump water than by working at the pumps BEP or Best Efficiency Point, which also means full speed.  So, adding a VFD will always cause more energy use per gallon produced and I am amazed at how many people think the opposite.

You are also correct that there is no more efficient way than controlling the pump with a regular pressure tank and pressure switch.  In this way the pump is either running at BEP or it is off and not using any power.  However, there are still many good reasons to use a Cycle Stop Valve and ways to make it as efficient as using a pressure tank only.  Offsetting the 6-9 times inrush current with each start is just one of many.

I will attach a curve from a 5HP, 25 GPM pump.  You can see this pump is most efficient (62%) at 25 GPM.  But you can also see that the horsepower drops from 5.5HP at 25 GPM to as little as 1.5HP at zero flow, even though the pump is running at full speed.  The CSV3A2T Cycle Stop Valve cannot close to less than 5 GPM no matter what, as this size pump needs at least 2 GPM for proper cooling.  At the 5 GPM minimum flow the pump is still 38% efficient and the horsepower has dropped from 5.5HP to 2.5HP with the simple restriction of the Cycle Stop Valve.  It is hard to wrap ones head around the physics because the amps dropping and the motor working easier as the pressure increases is counter intuitive.  But that is the way a centrifugal impeller works and always has.

Having said all that you can see that it doesn't matter so much how the pump is controlled, but the flow rate being used that determines the efficiency of the pump.  Therefore, even when using a CSV, keeping the irrigation zones or whatever is being used as close to BEP or 25 GPM will be the most efficient.  The Devil is always in the details.  So, how much of the time the pump will be working within 20% of BEP and how much of the time at other flow rates is important.  If say 80% of the zones are close to BEP, you may not notice any increase in energy consumption for the short times the pump is running at lower flow rates.  But if 80% of the time the pump will be operated at low flow rates, it will cost more per gallon to pump the water. 

There are also other things to consider.  The size of the pressure tank needed to reduce cycling to maximum factory standards can be expensive.  A 5HP is only allowed 100 cycles per day, and that is to make it last just barely past any warranty date.  As a matter of fact the number of cycles allowed and how a pump/motor is built makes pumps last an average of 7 years by design.  The fewer times a pump cycles the longer it will last.  Submersible pumps with Kingsbury type thrust bearings are completely frictionless when up and running.  Similar to a gas engine there is friction  until the pump gets up to speed and the lubricant/coolant starts flowing.  I have a pump in a stock well that hasn't shut off since it was installed in 1999.  I also know of other pumps that run continuously that have lasted longer than the 23 years that pump has lasted me so far. I expect to get another 23 years of use and I may not even be alive to see how long it lasted.

It takes a lot of energy to mine, manufacture, transport, and install a pump.  It is hard to quantify energy used, so lets figure costs.  Not considering inflation, lets say that pump cost $5K and has to be replaced 3 times in 21 years.  With an extra thousand bucks for a larger pressure tank and 15K worth of pumps, you will have spent 16K in 21 years more than if using a Cycle Stop Valve and having a pump last 2-3 decades.  The question now becomes, will adding $761.00 per year to equipment costs be offset by the efficiency of the system?

Starting a pump against an almost closed valve as with a CSV, is the best way to start a pump.  The mechanical soft start and soft stop created by the CSV, along with the elimination of repetitive cycling and other things makes everything from the pump and check valve to the pressure tank and switch last longer.

It is also an easy thing to test. The CSV uses the same pressure tank/pressure switch type system.  Simply run a month or two with the CSV and then tighten the adjustment bolt on the CSV a couple turns, which will disable it.  If after running another month or two without the CSV you are not convinced the CSV is superior, I will take it back and refund your money.  Been offering for 30 years and no one has ever sent one back.

[97044]

Interesting analysis and graph - it allows us to add some efficiency numbers.   But I am still not following.   

The power requirement of the pump pumping against 346 psi of head pressure is always going to be about 50% higher than pumping against 223psi of head pressure assuming the same flow (Q) (work = Q*dP).   So as a baseline in this configuration the CSV will require ~50% more power than using a pressure tank at all flows except the absolute maximum flow rate of the pump and the CSV is no longer regulating and providing back pressure.

This is further compounded by the pump efficiency issue at low flow.  The figure you are showing lets me put some numbers on that, at 5gpm the pump is consuming 2.5 HP or about 0.5hp per gallon pumped per minute [again excuse the terrible units].   At 25GPM the pump draws 5.5hp or 0.22hp/gallon/min.   So the pump ~250% more efficient at 25gpm than at  5gpm. 

So at flows close to BEP, the CSV would use 50% more energy to pump the same amount of water as a tank.  At the minimum flow (5gpm) the CSV would use 50% more energy due to back pressure on top of 250% more due to pump efficiency for 375% more energy in total.

I can see an argument that the CSV would cycle less (if the flow stays above 5gpm, below that the CSV would actually cycle more, unless you use a large pressure tank).  But the extra energy cost is pretty major.   

I appreciate the offer to try it out, but do you think the numbers will be far off what we are estimating here?

[Cary Austin]

At 25 GPM the CSV3A2T has only 4 PSI friction loss, which is so negligible you wouldn't even be able to measure a difference in power consumption of the pump.  The CSV only puts back pressure on the pump when less than 25 GPM is being used and back pressure is needed.

Those are not "terrible units".  Horsepower per gallon is the energy we pay for.

You are correct that at 5 GPM and 2.5HP, the pump is using almost five times as much energy per gallon as when pumping 25 GPM and using 5.5HP.  The efficiency goes up with the flow rate.  At 15 GPM the CSV is only 25% less efficient than a pressure tank. 

5.5HP for 25 gallons is 0.22HP per gallon.  This is exactly the same when using a CSV or just a pressure tank.

4.5HP for 20 gallons is 0.25HP per gallon.  This makes the CSV 12% less efficient that a pressure tank.

3.4HP for 15 gallons is 0.29HP per gallon.  This is 25% less efficient than a pressure tank.

3.5HP for 10 gallons is 0.35HP per gallon.  This is 62% less efficient than a pressure tank.

2.5HP.for 5 gallons is 0.50HP per gallon.  This is 227% less efficient than a pressure tank.

As you can see from these numbers it pays to stay close to 25 GPM.  But a little variation, even down to 15 GPM doesn't cost much extra. Having a few small uses less than 15 GPM is not going to add much to the electric bill either.  But if you have a lot of flows less than 15 GPM a smaller pump would be more efficient. Multiple pumps can come on as needed and off as needed by simply staggering the pressure switch settings. Multiple pumps makes it more efficient to have the right size pump for the job at hand.  Only one 86 gallon size pressure tank would be needed with the Cycle Stop Valve, even if using multiple pumps.

Without a Cycle Stop Valve a continuous flow rate of 15 GPM would cause a system with an 86 gallon size tank to cycle about 350 times a day.  It will take at least 4 of those large and expensive 86 gallon size tanks to get the cycles per day less than 100.  With 4 tank diaphragms cycling up and down a hundred times a day you can expect to replace them at least every 5 years or so.  Even if the pump survives 20 years, the costs of those extra three tanks up front and replacing all four every 5 years or so adds to whatever energy was used.  This is not even taking into account the multiple other problems associated with cycling the pump too much like water hammer, check valves, and pressure switches.

You have put your finger right on the problem of varying the flow rate of pumps.  I know you were not discussing VFD's here but VFD'S DO EXACTLY THE SAME TO THE HORSEPOWER NEEDED PER GALLON PRODUCED. 

If you do the math you will see that a CSV is as efficient as a VFD when controlling centrifugal pumps. But there is nothing more efficient than a regular pressure tank and pressure switch.

Cycling, using a regular pressure tank however causes many problems.  Adding a Cycle Stop valve can solve these problems, and can outweigh the reduced efficiency at low flow rates.




 

[Cary Austin]

Pumping into a cistern and using a couple of different size booster pumps fitted with Cycle Stop Valves can be even more efficient and would take a lot of cycling off the well pump either way.  Would be glad to discuss this further if you want to do the math?

[97044]

Thanks for the detailed discussion, I think we are agreed on the efficiency issues w.r.t to operating away from the pumps best efficiency point.

However I think we are talking past each other on the issue of back pressure.   To make sure I understand the process - in the scenario we are discussing, what is the pressure that the pump is working against at a flow of 5, 10, 15, 20, 25 gpm? 

At 25 GPM the CSV3A2T has only 4 PSI friction loss, which is so negligible you wouldn't even be able to measure a difference in power consumption of the pump.  The CSV only puts back pressure on the pump when less than 25 GPM is being used and back pressure is needed.
 

[Cary Austin]

OK well good question.  But back pressure has nothing to do with energy cost per gallon produced.  As you can see from the pump curve, as the back pressure on the pump goes up, the horsepower of the pump, which is the energy used, goes down.  This is counter intuitive and why many people have a hard time understanding how a valve works with a centrifugal pump.  It is the opposite of what you think as it would be inverted to use head or back pressure wrt efficiency or energy use.  The curve I posted shows the head in feet at all of those flow rates.  For example, at 15 GPM it shows 650' of head on the pump, which is the same as 281 PSI.  The curve shows head at any flow rate you want to pick, I wasn't trying to talk around it, I posted the curve.

Now why head or back pressure is important is to know if the pump will provide the head needed to lift the amount of water at least 400' and produce 50 PSI (115' of head) for the sprinklers.  We also need to know the max head and static level so we can make sure the pipe and the Cycle Stop Valve can handle the pressure.  The curve shows that pump will put out 25 GPM from 515' of head.  When the pump starts and the water level is at 400' it will produce 25 GPM at 50 PSI.  At this point there would only be 4 PSI more pressure on the inlet side of the CSV3A2T than the outlet pressure.  If the water level draws down more than 400' this pump will not be able to produce 25 GPM at 50 PSI.  Either the pressure or volume will reduce.  But if the well is strong and stays at 400' that pump with the CSV can produce 50 PSI constant as the flows vary from 25 GPM to as little as 5 GPM. 

When you are using 15 GPM for example, the pump will see 650' of head as noted earlier.  400' of that is needed to lift water 400', so the back pressure on the CSV will be 250' of head or 108 PSI.  Worst case back pressure for the CSV is when using 5 GPM.  There would be 740' of head or 320 PSI on the pump and well pipe.  But less the 400' from static, there would only be 147 PSI on the inlet of the CSV and the pipe prior to the CSV.  The CSV would maintain a constant 50 PSI on the discharge as the flows varied from 5 GPM to 25 GPM.

We also use the back pressure to determine the minimum flow through the CSV, but how that works is a whole other explanation.  Short answer is with 147 PSI in and 50 PSI out that makes a differential pressure across the CSV of 97 PSI, and the minimum flow will actually be more like 6 GPM instead of 5 GPM.  So, we always figure back pressure but it doesn't change the horsepower curve.

[97044]

OK well good question.  But back pressure has nothing to do with energy cost per gallon produced.  As you can see from the pump curve, as the back pressure on the pump goes up, the horsepower of the pump, which is the energy used, goes down.

This makes no physical sense.   It takes more energy to move a unit mass of water against a higher back pressure than it does against lower back pressure.   This is the whole premise of the work equation: work = flow*delta_pressure.  The horse power of the pump goes down not because of higher pressure, but because the pump is moving less water and therefor doing less work.  As we demonstrated earlier the amount of energy needed to move the same amount of water goes way up as the head pressure increases.

To convince your self of this, think about it in terms of vertical head instead of pressure (the pump doesn't know the difference), and let's use SI so that the units make sense.  In this case we are then thinking about moving a certain amount of mass per second up a certain distance against gravity.     To make it obvious, let's go back to the 5 GPM case:

With the CSV the pump sees 740' of head, or 225m.   We are pumping at 5 gpm, or 0.32 kg/s of water.   So the pump is lifting 0.32kg of water 225m every second.   The best case scenario, the power requirement is  (Height*mass*gravity):  225*0.32*9.8 =  705J/s or 705W or about 0.95hp.

Without the CSV valve, the pump sees 515' of head or 157m.  The pump is still providing 5 gpm average by cycling on and off.   So the pump is lifting 0.32kg of water 157m every second.    Now the power requirement is 157*0.32*9.8 = 492 J/2 or 492W or 0.64hp. 

Regardless of pump efficiency, the increase in head pressure from the CSV increases the load on the pump from 0.64hp to 0.95hp.    In reality, because pump efficiency also goes down at higher head pressures, the situation is way worse - the pump actually needs 2.5hp at 5gpm, so the overall efficiency is 0.9/2.5 = 36%.

The follow on question, where is that 0.3hp going?  It has to be dissipated in the CSV. 

So yes, back pressure absolutely changes the amount of energy per gallon produced.  This is fundamental to Bernoulli's equation:
https://en.wikipedia.org/wiki/Pump#Pumping_power

This may explain some of the confusion about Variable Speed Pumps.

[97044]

Pumping into a cistern and using a couple of different size booster pumps fitted with Cycle Stop Valves can be even more efficient and would take a lot of cycling off the well pump either way.  Would be glad to discuss this further if you want to do the math?

This is a good suggestion, and maybe what I am leaning towards now. 

[Cary Austin]

Hydraulic HP = Head (ft) x Flow Rate (gpm) x (Specific Gravity)
                                                         3956

This is the formula for figuring Brake Horse Power.  But head is irrelevant when the curve already shows the horsepower. 

You are comparing the CSV working at 5 GPM and the system without a CSV working at 25 GPM, which makes no sense.  When running the CSV at 25 GPM the pump will only see 4 PSI or 9' of head more than if there was no CSV, which is 524' instead of 515'.  Do the math or just look at the curve and you won't be able to see any appreciable difference in horsepower with only 9' of head difference.  I can even add a bypass that will keep the CSV from adding 9' of head when wide open if needed.  But it is not worth the effort.

When using 5 GPM with the CSV the curve shows 2.5 HP, or 0.5HP per gallon, no need to do the math. We already know that without a CSV the pump will only be using 0.22HP per gallon, but will be cycling itself to death while that is happening.  Actually it will be a little more than 0.22HP per gallon if you figure the number of cycles and add the inrush currents for each cycle.

I am not confused about VFD's but most people are.  Even though the benefits of using a Cycle Stop Valve can far outweigh any loss of efficiency, of course restricting flow from a pump is less efficient.  But it is the same thing when using a VFD.  I wish others would do the math on VFD's and see that reducing the pumps speed also increases the horsepower per gallon used and is the same as restriction from a CSV.

[Cary Austin]

When pumping into a cistern the well pump is always pumping at BEP.  Only the booster pumps will be working from one end of the curve to the other.  But the main job of lifting 400' can be done with a smaller pump and use less energy as it will not be making the extra 50 PSI needed.

[Cary Austin]

A submersible in the cistern can be more efficient than an end suction centrifugal.  Here is a curve for a pump that a CSV will make work from 1 GPM at 0.4HP to 25 GPM at 1.44HP.

The pump in the well will now be able to fill the cistern at 30 GPM using the same 5.5 HP energy it was using for 25 GPM.

[97044]

You previously stated that 'But back pressure has nothing to do with energy cost per gallon produced.'  Your own equation (and the example we have hashed through) clearly shows that to be false.  Increasing back pressure = increased energy per gallon produced.

When using a pressure tank to provide 5gpm, the pump is running at 25gpm, but only for 20% of the time, the pump never runs at 5gpm.

Since you keep bringing up VFDs or VSDs, isn't the situation the same as with a tank?    A constant pressure VFD is only pumping against a back pressure of 50psi (assuming that is the setpoint) + the lift pressure.    So even neglecting the impact of being closer to BEP at lower RPM,  just the reduction in back pressure nets the same 35% reduction in energy as with a tank.    If you include the increase in pump efficiency by operating closer to BEP, then the VFD looks even better, no?

If you have physics that shows otherwise, I would be keen to see it.

Hydraulic HP = Head (ft) x Flow Rate (gpm) x (Specific Gravity)
                                                         3956

This is the formula for figuring Brake Horse Power.  But head is irrelevant when the curve already shows the horsepower. 

You are comparing the CSV working at 5 GPM and the system without a CSV working at 25 GPM, which makes no sense.  When running the CSV at 25 GPM the pump will only see 4 PSI or 9' of head more than if there was no CSV, which is 524' instead of 515'.  Do the math or just look at the curve and you won't be able to see any appreciable difference in horsepower with only 9' of head difference.  I can even add a bypass that will keep the CSV from adding 9' of head when wide open if needed.  But it is not worth the effort.

When using 5 GPM with the CSV the curve shows 2.5 HP, or 0.5HP per gallon, no need to do the math. We already know that without a CSV the pump will only be using 0.22HP per gallon, but will be cycling itself to death while that is happening.  Actually it will be a little more than 0.22HP per gallon if you figure the number of cycles and add the inrush currents for each cycle.

I am not confused about VFD's but most people are.  Even though the benefits of using a Cycle Stop Valve can far outweigh any loss of efficiency, of course restricting flow from a pump is less efficient.  But it is the same thing when using a VFD.  I wish others would do the math on VFD's and see that reducing the pumps speed also increases the horsepower per gallon used and is the same as restriction from a CSV.