Pumping uphill

[NHmaster3015]

Hunter Technical Manual 2002

Note the part that says plastic IPS not glue

Now I'm outta here because as Nukeman says we need to compare apples to apples and nobody seems to want to do that so the entire discussion is pointless. Especially when simple laws pf physics are ignored.

 

[Valveman]

If you had large enough pipe to keep the head at 133’, the pump in your picture would be much more efficient because it is larger in diameter. Of course you couldn’t afford the starting current for a pump that size for just 500 gallons. However, after start up, that pump would put 500 gallons out for much less than a smaller pump at the same head.

Moving the numbers up from 1 to 10 GPM and 5 to 50 GPM gets us in a range where we can find pumps that fit the design requirements. However, it is the difference in friction loss that causes the difference in power consumption, not the rate of flow. If the head where the same, a 50 GPM pump would just fill the tank 5 times faster than a 10 GPM pump, but would use exactly the same amount of energy per gallon. It would take 0.56 HP to pump 10 GPM, or 2.8 HP to pump 50 GPM at the same head. Five times the flow is exactly five times the bhp.

I am not arguing the fact that lower head requires less energy. However, while it would be easy to find a pump that would do 10 GPM at 0.56 HP, or 50 GPM at 2.8 HP, it is not possible to find a pump that will do 1 GPM at 0.056 HP.

It is possible to find a pump that would do 5 GPM at 0.419 HP. Restricted or slowed to 1 GPM would reduce bhp to about 0.35 HP. Because a ½ HP pump is the smallest available for this type of application, it would still use less energy to let the tank fill in 100 minutes instead of 500 minutes. If anybody knows where to get a 1/20th HP pump that will do 133’ of head, I am all ears.

 

[Sponsor]

 

[Ballvalve]

Perhaps one of those DC solar pumps at 1,000$ would come close and with another 2 grand in panels, he would be up and going. Then, he should use 1" black poly in 600' lengths with exterior couplings or butt welds so as not to restrict flow.

 

[Valveman]

Nobody is ignoring the laws of physics. But if you are going to use the laws of physics you have to add in all the variables. That includes efficiency loss at partial load, parasitic losses, and working within the characteristics of a real pump, not an imaginary one. You can’t compare apples to oranpearlopes, because they don’t exist. I have no argument with Nukeman’s numbers, just that he is specifying a pump that doesn’t exist.

Also, 1” Sch 40 and 200 class pipe is IPS, or Iron Pipe Size. Even if it where bell and gasket pipe instead of glued, gasket pipe would have the same friction loss as belled ends.

 

[NHmaster3015]

No it doesn't. You chart refers to threaded plastic, not glued. Check Charlotte pipe's web page and the friction loss for PVC glue fittings. Couplings ain't there, nor are they for any PVC glue manufacturers couplings

The original discussion was operating the same pump at different pressures. Never mind what pump. The"what pump" is where the whole discussion went south.

 

[Ballvalve]

That chart says 1" steel coupling has 1 foot of loss VS 3 foot of loss in plastic. That seems incorrect. Steel joints should have all sorts of turbulence.

 

[Gary Slusser]

valveman: I agree with that. I'm talking about what the requirements should be. If you can't find a pump small enough to do the job, then you choose the next best thing. Pick the smallest pump that will do the job and weigh cost, quality, parts availability, etc.

This is a homeowner in Utah, maybe at 8500'+/- asking about a pump and most of us here that answer those questions suggest the most common off the shelf pump etc..

That usually is the best choice based on your criteria of cost, quality, parts availability, etc.. We don't get into minutia engineering and go off into nuclear power plant size pumps etc..

What I say still holds true:

pump 1: designed for 1 gpm at 133'
pump 2: designed for 5 gpm at 199'

Pump 1 will use less energy to do the job. Simple as that.

I take you to mean that his electric bill will be the same or less than the larger pump because that is how he will judge all this.

How about you find, propose and give us a name brand, model number and voltage required of those pumps based on your criteria above, I and others here can't seem to find them. And does his elevation have anything to do with your selection of hp or not?

Gary started out saying that a huge pump and 2" poly would be needed for this job and that is simply not true. 1/2 hp is way more than enough. Remember also that we aren't limited to pumps designed for wells. Maybe a well pump would be best for the job, but maybe not.

A 1/2 hp at sea level but is questionable at 5000+ feet, no?

Yes I didn't pay enough attention to the part about a dribble into an atmospheric tank and I was thinking of him not wanting to run a pump for 24 hours to refill the tank or the cost of the 2500' of large cable etc. etc..

So I suggested the slickest water line material known to man (PE) requiring the minimal number of fittings, which meets your cost criteria and his ease and time spent in installation while limiting the possibility of underground joint leaks (only 2 joints underground in the 2500'), and any expansion/contraction problems with the use of sch 40 PVC, that most homeowners usually choose because they aren't aware of PE pipe but...

Although the 1.5 hp I mentioned could be a very good choice, the 5 hp I mentioned is probably overkill; unless he now or later decided to forgo the pump his buried storage tank at the house now requires and run water from the spring to his faucets in the house for say a peak demand of like 10-15 gpm at 40/60 psi so he could get rid of the water quality and maintenance problems with a cistern. Which brought me to the 2"-2.5" line size I mentioned.

Problem is he hasn't come back to discuss anything like simplification of his system or possible future water use needs or, how many gpm the spring produces and how big a tank he has at the spring (he mentions use of his portable 500 gal tank but doesn't say he fills it, etc..

 

[Zl700]

Oh silly plumbers

IPS means "Iron Pipe Size" that doesnt mean that it has to have threads!!!

Such as black poly pipe that comes in IPS or CTS (copper tube size)-100#, 160# & 200# WST (Water service tubing)

Who ever heard of threaded 160# IPS or CTS black poly coils???

So glad you now know couplings do have a friction loss.

 

[Valveman]

That chart says 1" steel coupling has 1 foot of loss VS 3 foot of loss in plastic. That seems incorrect. Steel joints should have all sorts of turbulence.

I agree but, there is more open space inside a steel coupling than a PVC one. And maybe they figured everybody goes crazy with the glue? I don't see anything about threaded.

 

[Nukeman]

valveman: I agree with what you are saying. As I have said, the smaller flow should use the same/less energy. If the head is the same and the efficiency is the same, then both cases will use the same energy. However, keeping the rest of the system the same, there will always be more head at the higher flowrate (which means more energy used). Sure, you can increase the pipe size, etc. to counteract, but this costs $$$. The big one is the power requirement. Going up just one or two sizes on your power feed makes a big difference when you are talking about 2500 ft of cable. That is why I suggested that a small pump + solar might be a good option.

If you had a small recirc pump that could do the small flow and not the head, you could have a couple pumping stations along the way to make it happen. Again, it may/may not be the best solution, but would be another option.

The fact is that we don't know enough about the problem to know what is best for the application. Right now, we can only guess how often the transfer needs to be made and what we have for power. Once that is known, we could have moved on to selecting a pipe type/size, pump size, electrical cable size, etc. Then it is a matter of trading one thing for another. For instance, poly might be more expensive than PVC, but the head might be reduced enough to go 1 step down on pump size and still meet the specs. If the pump is small and the cost of running power cable 2500' is large, solar might be an option.

It would be nice if the OP came back. This is a good design project. It would be interesting to see what everyone comes up with and hopefully combine the experience that will make for a good design. Like anything in engineering, you trade one thing for another. Pick the pipe, determine the head, pick a pump, then pick the cable. Then you might find you are just above say a 1/2 hp pump. You could then adjust the piping some to get into the sweet spot of the 1/2 hp pump you selected. Again, you would have to weigh the pump price difference/operating cost difference with piping/cable cost difference and see what makes the most sense.

The coupling thing is interesting. One would think that they were talking about screw-type fittings from the significant loss, but it also mentions copper (which is normally a soldered sleeve). Hmm.. unless they are talking about larger copper and using brass/threads to make the connection. Also the loss for steel is listed at 1/3 to 1/2 that of plastic/copper. Seems strange as steel would normally be done with screw-type fittings... I agree that a correctly glued PVC coupling should have almost no loss. That chart is saying that a coupling is like having a 45* elbow in there. I can't believe that one. I did take a look online and a few places show a loss for PVC couplings, but they are talking about "internal couplings." Like this:

Perhaps this is what they are talking about in this chart.

Having a coupling "more open" in the middle actually adds to the loss. Whenever you change the flow area (expansion/contraction), there is a loss and that loss it related to the ratio of the flow areas.

 

[Nukeman]

This is a homeowner in Utah, maybe at 8500'+/- asking about a pump and most of us here that answer those questions suggest the most common off the shelf pump etc..

Elevation has nothing to do with it. The atmospheric pressure at the spring is the same as the house (with the exception of the 120' elevation change, which is insignificant). The pressures cancel each other out (DP).

That usually is the best choice based on your criteria of cost, quality, parts availability, etc.. We don't get into minutia engineering and go off into nuclear power plant size pumps etc..

I take you to mean that his electric bill will be the same or less than the larger pump because that is how he will judge all this.

Yes. A properly sized pump at slower flows will cost the same or less in electricity than a larger one given everything else being the same. The additional advantages of a smaller pump were also mentioned (wiring cost, piping cost, etc.). You may find a condition where the larger pump might be cheaper to operate, but that would be with higher capital costs (larger piping etc.). Then you are not comparing one system directly with the other. If we had more info, we could investigate capital/operating costs and see what option is better over the time he is expected to live there.

How about you find, propose and give us a name brand, model number and voltage required of those pumps based on your criteria above, I and others here can't seem to find them. And does his elevation have anything to do with your selection of hp or not?

Okay. How about a DC submesible pump? Maybe something like this Shurflo 9300 pump. Runs off two 80W solar panels. 1.65 GPM at 140' head, 1.63 GPM at 160' head. 3.1A at 30V means that it uses 90W (under 1/8th HP). The panels may be cheaper than running at that wire. If you did have power there and didn't want to go solar, just use a transformer to step down your 120v to 24v. This pump looks like it is arond 50% efficent at these conditions.

http://www.solar-catalog.com/pumps_submersible.html

You could also go with another style of positive displacement pump, but most are designed for higher viscosity/lubricating fluids (oil, gasoline, etc.). Elevation does not factor into the pump sizing. You could also go with a hydraulic ram (already has been mentioned) depending if the elevation continues to fall beyond the spring and if it does, it is part of his property. We need more info for a proper pump selection.

A 1/2 hp at sea level but is questionable at 5000+ feet, no?

Again, this does not matter. You might consider the pressure difference if he was pumping across a large elevation difference (like say he had to pump 2500' vertically), but even in this case, it doesn't really matter. Remember that a liquid (like water) is roughly 1000x more dense than air/gases. That means for every meter water head, the atmospheric pressure change would be equivalent to 1 mm of head (tiny in the scheme of things).

Yes I didn't pay enough attention to the part about a dribble into an atmospheric tank and I was thinking of him not wanting to run a pump for 24 hours to refill the tank or the cost of the 2500' of large cable etc. etc..

So I suggested the slickest water line material known to man (PE) requiring the minimal number of fittings, which meets your cost criteria and his ease and time spent in installation while limiting the possibility of underground joint leaks (only 2 joints underground in the 2500'), and any expansion/contraction problems with the use of sch 40 PVC, that most homeowners usually choose because they aren't aware of PE pipe but...

Nothing wrong with PE and it may be the better choice here. Certainly save time without gluing all those couplers. Just don't need to use 2" poly. 1" or smaller would be fine.

Although the 1.5 hp I mentioned could be a very good choice, the 5 hp I mentioned is probably overkill; unless he now or later decided to forgo the pump his buried storage tank at the house now requires and run water from the spring to his faucets in the house for say a peak demand of like 10-15 gpm at 40/60 psi so he could get rid of the water quality and maintenance problems with a cistern. Which brought me to the 2"-2.5" line size I mentioned.

At 230v:

For 1.5 HP, you would need 2 ga. copper to make 2500' ($$)
For 5 HP, you would need 3/0 copper to make 2500' ($$$$$)

At say 1/2 hp, 4 ga would do it. Might be able to even go 6 ga with a smaller pump.

Problem is he hasn't come back to discuss anything like simplification of his system or possible future water use needs or, how many gpm the spring produces and how big a tank he has at the spring (he mentions use of his portable 500 gal tank but doesn't say he fills it, etc..

Agreed. Not enough info here. We can only guess at the specs. I do think it would be a good exercise to work though the whole design and take the experience of the people on here and get a really good system setup.

 

[Valveman]

I hadn’t thought of a DC motor and a transformer. Those solar powered pumps would give the head required, even at that small of a flow rate. Of course I don’t think those little solar pumps are very long lasting, so you might be paying more for equipment than you would save in energy.

A few fractional horsepower circulator pumps staggered every 30’ of lift along the way is another interesting idea. Might be more complicated than it is worth though.

I can’t believe the OP hasn’t come back. Where else could he have gotten so many interesting ideas and advice from a Nuclear Engineer? Oh Well!! I enjoyed the discussion! Thanks Everyone!!

 

[Nukeman]

See, discussions like this are really great because different people have experience with different things. You collect the ideas and take the best of them and combine them together. I think that between all of us, we could have come up with a really great system.

You are right in that everything has to be weighed. Upfront cost, cost to maintain, cost to operate, etc. Other things to consider would be future parts availability and ease of service/rebuilding. Since power and pipe would be run the whole distance (assuming solar wasn't used), pumping stations would be easy to incorporate, but again, you would have to weigh the cost of 3 or 4 pumps (plus maintaining them) that could do 40'-50' of head vs. one larger pump that could do the whole thing.

A 1/2 HP well pump might be the best overall because you know that you will be able to replace it at any time, plus parts should be easy to get, etc. I do like the idea of solar for something like this. I don't think the cost would be much higher (if any) than laying 2500' of cable especially if you are paying labor to lay that 2500'. Solar would also be good for this because it sounds like it is something that doesn't have to run all of the time. When the sun is hitting the panels, it can be pumping away and then can shut off at night or when the tank is full.

Something like this would be great for a project for someone starting an engineering program in college or could be a great senior design project for high school. This is a real world problem here. You have some requirements that need to be met and then you go out and design something that'll do the job. Even better is seeing that design built and seeing it work like it is supposed to. Lots of things to consider here and many of these things are compounded by the fact that the distances are so long. If you only had to run 25' or 100', going to larger pipe, cable, etc. would not be a big additional cost, but when you are talking about 2500', those costs add up fast. Not to mention, you have to start with larger everything to begin with (flow resistance, electrical resistance, etc.) due to the long runs.

Another intersting thing to think about is the electrical resistance loss. The cable size that I mentioned before is for a 5% drop in voltage over the distance. Assuming 230v supply, this works out to 11.5 volts. If we then multiply this by the pump current, we get the power lost due to resistance. At 2 amps, this is 23W and at 20 amps it is 230W (about 1/3 HP). This all works into the system efficiency. So, we have: power coming in, line loss, losses in pump motor, losses in the pump impeller, etc. All would need to be taken into account.

Then you might have:
- System A: capital cost (materials/labor), operating cost ($/month or $/1000 gallons pumped), cost to maintain (could be added to operating costs), etc.
- System B: ....
- System C: ....
- System D: ....

Then you just select the 'best' option. Usually total cost (over the expected time to use the system) is the most important, but there may be other factors. For instance, it might be cheapest with a cheapo pump that needs to be rebuilt/replaced every year, but maybe he does not want to mess with that and is willing to pay a little more for something that might go 10 years without rebuilding. Another example might be solar. Say the panels cost more than buying cable. It still might be worth it to go with solar since laying 2500' of cable is a ton of work (I get sore just thinking about it. ).

It has been a great discussion. I do hope he comes back so we can complete this instead of leaving it as something half done with only assumptions to go on. He might come back expecting to see one or two responses and ends up with a giant thread instead.

 

[Ballvalve]

Franklin allows for 6 ga. copper for 2500'. If he used a 3 wire pump with the control box at the top of the hilll, and a 1/3 hp motor [found some online by franklin] i would guess 8 gauuge might suffice. These motors will tolerate a lot of voltage drop.

Hot tip for you guys wanting to play with solar ; Amazonia has a pack of 4] 15 watt panels with controller and inverter and rack for 289$ delivered. Thats beyond cheap. and 84 reviews say they hold up pretty well. Was at 600 bucks. Now the guys numbers work with a little dc pump even if you replace it every year.

Gary, get a physics class. your submersible will work the same on Mt. everest. Actually I think the suction lift for a shallow well pump increases nicely with elevation rise.

 

[Valveman]

Actually I think the suction lift for a shallow well pump increases nicely with elevation rise.

Actually I think it is just the opposite. I also think elevation effects submersibles as well.

 

[Ballvalve]

Yes I see 17' of suction lift at 6000'. Better move to death valley to get 30 of lift.

Good question about submersibles.... However, I have never seen a submersible pump curve mention elevations above sea level. That effect must be too small to mention.... but I suppose it means that a submersible pump set at 1,000' [below sea level] with water at 900' would move more water than one at 30' deep. [?]

Perhaps with PUSHING water with a submersible, lower atmospheric pressure [higher el above sea level] would give an easier lift, i.e contrary to suction lift ?

 

[Valveman]

Perhaps with PUSHING water with a submersible, lower atmospheric pressure [higher el above sea level] would give an easier lift, i.e contrary to suction lift ?

Interesting!! Probably right but woudn't be as critical as suction lift. Otherwise with submersibles, I think it only has to do with the amount of submergence. You can have less water over the intake at sea level than at 5,000'.

 

[Nukeman]

You are looking at NPSH.

http://www.engineeringtoolbox.com/npsh-net-positive-suction-head-d_634.html

In the case of the submersible, it would be similar to the case of a pump located below a tank. With everything else the same, NPSH changes with vapor pressure. This changes with temperature as well as elevation (reason why water boils at a lower temp at higher elevations).

NPSH is a big deal for nuke plants, especially in an accident. The system may normally be at 2250 psi and 625F. You have a break and a crap load (technical term ) of steam/water droplets come flying out of the break. The containment building is sealed up, so the pressure/temperature increases. The water that ends up on the floor goes to a tank and is sucked up and pumped back into the system to keep the core covered. As the water heats up, the NPSH starts to fall rapidly. Things need to be designed so that the pumps can still pull the water from the tanks under the worst conditions.

NPSH will drop with higher elevations no matter if the pump is above or bellow the water. The reason is that the NPSH only considers what is going on on the suction side of the pump. As long as your pump can cover the NPSH, it doesn't matter what the atmospheric pressure is doing.

 

[Gary Slusser]

Solar and a small Shurflo pump is a bad choice. Especially in winter or cloudy weather. I live most of the year off grid using solar and a 3-5 gpm at 35 psi IIRC Shurflo pump in my motor home. And it doesn't meet Nukeman's requirements f cost, quality etc.. You should check if the pump is rated continuous duty.

I'm not going to go look it up but from memory, pump charts are done at sea level and as you go up in elevation atmospheric pressure decreases and the out put of the pump decreases. IIRC, you are to go up to the next HP I think above 5000'. Or for every 1000' above 5000'.

That's why a jet pump at higher altitudes can't lift the same 25' it does at sea level. Subs can't move as many gpm at higher altitudes as they do at sea level. Ballvalve, maybe a pump sizing book instead of physics.

 

[Nukeman]

You can't know that solar is a bad choice, because there is not enough info. How do we know that he doesn't use this water for irrigation and only uses it in spring/summer? Maybe it is water for his house, but maybe he has a winter home and is away during the winter? Who knows? Again, all options should be considered. Running 2500' of cable is expensive and there is a good amount of line loss that goes with it (especially at higher currents). We are just guessing at this point at what the real requirments are.

You wanted to know of a pump that could do the low flow and high head (you didn't think such a thing existed) and I gave you an option. There are other options too. We don't know what is best because we don't know all of the requirements.

In terms of elevation, the following happens with basically any pump:

- NPSH decreases with elevation
- flow and head decrease in gasoline powered pumps (reduced engine performance)

For an electric pump with both sides open to atmospheric pressure, the flow will not change with altitude. See, the suction head goes down with increased altitude, but the discharge head also goes down the same amount. The pump does not know it is working at a different altitude. Your charts are looking at a fixed discharge head, which would be true for pumping into a sealed/pressurized system. You have to remember that this setup is not quite like the typical well setup you are used to seeing as it is a non-pressurized system.