Insufficent watefall pump size !!! What size should i use

[mas985]

Unfortunately, most medium and high high pumps do not publish flow rates that high in the head curves so it is uncertain if they could achieve that. Theoretically, they could with low head but the pump itself adds a certain amount of head loss so even if the plumbing were zero head loss, there is still head loss in the pump and the pump would max out at some flow rate. I suspect this is the reason that most pumps don't show flow rates above 180 GPM. The 3 HP Northstar goes to 180 GPM but at 60 feet of head. So it is not clear what happens to many of these pumps above the published levels and extrapolation can be erroneous.

The Pentair waterfall pumps are very low head waterfall pumps. The Hayward Tristars have good flow at higher head values but the specs start at 40 ft of head so it is difficult to say what happens at lower head. Discussion with the pump manufacture may help but I suspect that they don't even measure the performance beyond what is published in the head curves.

Another option is to use two matched parallel pumps. You can get away with 2 lower HP pumps while doubling your flow rate for a given head loss. A detailed analysis of this type of configuration is a bit tricky but can be done. The pumps see the same head loss which is determined by their combined flow rates. One easy approximation is to assume a single pump with plumbing of 2.3" diameter or 3.6x the equivalent length and then double the resulting flow rates. So a single pump with 2.3" plumbing will have the same head loss as a parallel pump system with 3" plumbing. So for example:

[EDIT] Corrected these numbers such that the plumbing would yield 26 feet of head at 175 GPM.

A single 3/4 HP Whisperflo with 2.3" plumbing yields 94 GPM @ 30 ft of head so two pumps with 3" plumbing would give you 188 GPM at with the same 30 ft. of head.

A single 1 HP Whisperflo with 2.3" plumbing yields 108 GPM @ 37 ft of head so two pumps with 3" plumbing would give you 216 GPM at with the same 37 ft. of head.

The pumps are still running at a low head loss but it is within the operating range of the pump this way. The pumps will run quieter than a single high HP pump and with less energy use (2x1.3kw vs 3.1 kw for 3HP). The downside of course is the cost of two pumps. However, this technique could also be applied to lower cost less efficient pumps as well such as the Pentair Superflo.



BTW 3" 90's are closer to 8 feet of straight pipe.

 

[carlos31]

Thank you very much JasonLion and mas985.

First, let me provide this link where I obtained my information for calculating head in case anyone needs it.

http://www.engineeringtoolbox.com/resis ... d_192.html

Second, the pump curve for the Whisperflo pump JasonLion suggested indicates 160 gpm at 50ft of head. Does anyone know what will happen if I use that pump at 26ft of head? Will it simply not work unless I increase the head by constricting the line with a valve or will it operate at the max 160gpm indicated on the pump curve, or will it move water beyond 160gpm? :?:

This idea of not enough dynamic head is very perplexing :? I assume that this is some point where the pump simply will not move more water even though in theory it should be easier for the pump to do so. After all, if it did move more water at lower head why wouldn't that be on the curve? mas985 touched on this very question and provided guidance but not a silver bullet answer. I hate it when there isn't a silver bullet answer! :-D

Third, I also found the Northstar SP4030NS and the table in the manual indicates 206gpm at 40ft head. I have noticed that Northstar is NOT everyone's favorite pump but this one may be the winner of the pump curves for my situation. I also looked at 5hp pumps and light commercial pumps online until I saw the prices.

Fourth and finally, thank you mas985 for the great idea of using parallel pumps. If I really want to move serious water; say maybe 200gpm to a waterfall and another 40gpm to a slide then that may be the best way to go. The price of two "normal" pumps doesn't look so bad after you have checked out the light commercial and 5hp pumps.

 

[JasonLion]

The further to either side of the pump curve you go the more you get outside the high efficiency band, which is normally in the middle of the pump curve. At the far left of the curve the pump will start to cavitate, towards the far right I am less clear on what happens.

I hadn't thought of using two pumps. The dual 3/4 HP configuration mas985 suggested has the advantage of saving some electricity, since the pumps will be operating closer to their ideal zone.

 

[mas985]

Carlos31,

I use that site often for my calculations but the tables you are using are for hot water metal pipes. You need to use the PVC version which is located here. I used much of this information to write this sticky.

Second, the pump curve for the Whisperflo pump JasonLion suggested indicates 160 gpm at 50ft of head. Does anyone know what will happen if I use that pump at 26ft of head?

It is likely to top out at some flow rate slightly above 160 gpm but again when you operate at such high flow rates, the pumps tend to run very hot and very noisy. When you get far above the best efficiency point (BEP), this is what happens. The BEP for that pump is 123 GPM at 73 feet of head. This is where the pumping efficiency is the greatest. Above that point more of the electrical energy is converted to heat. At 160 GPM, you will be running at 30% over BEP which I would not recommend. Most pools are within +- 10% of the BEP point which is about as good as you can get.

After all, if it did move more water at lower head why wouldn't that be on the curve?

Probably because the manufacture doesn't recommend operating in that region either.

Third, I also found the Northstar SP4030NS and the table in the manual indicates 206gpm at 40ft head. I have noticed that Northstar is NOT everyone's favorite pump but this one may be the winner of the pump curves for my situation. I also looked at 5hp pumps and light commercial pumps online until I saw the prices.

I have two Northstars, one for the pool and one for my spa and I have been quite happy with them and have not seen any problems that are often quoted on pool forums. That is not to say that they don't have problems. I suspect that most of the problems are caused by poorly designed plumbing. Northstars have unusually high service factors (1.85) which means that pool builders are probably putting them on plumbing which cannot support them. This forces the pump to run outside of the most efficient regions of the head curve and thus probably have more stress on the pump causing premature failure.

Having said that, the 3 HP Northstar's BEP is 149 GPM at 74 feet of head. At 206 GPM, that is close to 40% over BEP again I would not recommended that.

Although costly, two pumps will probably be your best bet and it could give flexibility with the flow rates. If you put one way valves on the output of each pump, then you could run only one pump at a time if you want a different action for your waterfall. Running one pump would give you about 104 GPM at 15 feet of head for the 3/4 HP Whisperflo vs 188 GPM at 30 feet of head for both pumps. So not quite half the flow rate.

 

[carlos31]

achieving some understanding!

Thank you for getting me onto the correct engineering graph and for explaining why I need to be in the efficiency sweet spot of the pump curve.

I recomputed the head for 200gpm and arrived at about 34ft (9.9 suction and 23.8 return). This appears to put me in good shape for either the dual 3/4hp or 1hp Whispeflo pumps. I expect to get a 4x160 for my main pump, so I could use it to try different flow rates on the waterfall plumbing before its permanent installation. Based on that information I can go 3/4hp or 1hp. I guess I need to talk to the plaster guy to check if that is even feasible. I may need to install it, wait until the plaster is cured, uninstall it and move it to the waterfall plumbing, test it, then move it back again. Sounds like a pain but its probably better than guessing.

The only wildcard would be the 100 ft long 1.5 inch line to the waterslide which is parallel to the waterfall return. The slide can handle up to 40gpm although it may not need that much flow. Will the opening and closing of this valve make a huge difference in the head? How do I compute for this additional variable? Other than this I feel like I am ready to go thanks to your help. I have learned a ton!

Also, please confirm that I should put a valve on the waterfall return side so that I can increase the head if necessary? Is this a situation where you might advise I install a flow meter or pressure gauge? If so, where do I place them?

Thank you.

 

[mas985]

So that is an equivalent length of pipe of 115 ft for the suction and 185 feet for the return (+8' for vertical rise)?

Using those numbers, two 3/4 HP Whisperflo's would give you about 187 GPM at 31 feet of head. Two 1 HP Whisperflo's will give you about 219 GPM at 38 ft.

You have a couple of options when plumbing this. To give you the most flexibility in adjusting flow rates, I would put valves on both the waterfall line and slide line. Because the waterfall has a larger line, it will have the least resistance to flow and lowest head loss. When the slide is turned on, the total head loss seen by the pump will drop slightly and flow rate will increase slightly since the water has two paths it can now flow through. However, the waterfall flow may actually drop a bit because it is sharing water with the water fall. If you also place one way valves on the pumps so that you can turn one pump off, you would have the ability to run only the slide if desired. I would not run both pumps for the slide only. Also, adding automation to the valves would simplify things a bit.

Flow meters are nice to have but not really necessary. Unless you have a specific reason as to why you need it, I would skip it. You can always add one later. You will know how to adjust the valves by simply watching the water flow for both the slide and waterfall. As for circulation, a better metric than flow rate and turnover is how the water looks.

Also, I assume you plan on bypassing the filter and any heater for this loop as the flow rates will overwelm the filter and perhaps damage it. It would also add head loss that you need to avoid.

 

[carlos31]

waterfall and slide

Great advice. The waterfall/slide plumbing is a standalone system; no heater or filter comes into play. When you say "one way valve", do you mean a check valve? I assume a simple ball valve is considered a two way valve and I will certainly will use them on the return side of both return lines.

I may not be able to use the slide without the waterfall but I don't think that is a big deal. I presume the output of a single 1hp Whisperflo will exceed the max slide output of 40gpm and could break something in the slide.

 

[mas985]

Yes, I meant a check valve and you are correct in that the 1 HP will be well over 40 GPM.

 

[carlos31]

Thanks for all your help. I appreciate it. It will be about three to four months before my pool is finished. I will post some pics then.

 

[carlos31]

I am adding these PMs between mas985 and myself to this post in case others are interested in the information. The subject is using parallel pumps to supply a large water feature.

Message subject: Re: Insufficent watefall pump size !!! What size should i use Sent at: Thu Jul 10, 2008 7:13 pm
Thanks, I will move this to the post.

I understand your explanation. I guess my real problem now is that I have no reference for how much water is required for a gushing waterfall. The few waterfalls I have seen which are similar to my own have been lackluster performers and all used a single 2.5 hp pump. So I know what not to do; I just don't know what to do.


View message

mas985
Message subject: Re: Insufficent watefall pump size !!! What size should i use Sent at: Thu Jul 10, 2008 5:55 pm
carlos31 wrote:
thanks for trying to explain this to me, but I am still confused.

You wrote "A single 1 HP Whisperflo with 2.3" plumbing yields 108 GPM @ 37 ft of head so two pumps with 3" plumbing would give you 216 GPM at with the same 37 ft. of head."

Should I expect 216gpm actual output OR is this only theoretical because using two pumps will increase the dynamic head far beyond the 37ft when using one pump?

Sorry, I am not catching on quickly here!


If you re-read the example I gave carefully, you will not that two things are changing. First, going from one pump to two pumps AND the plumbing size changed from 2.3" to 3". This was so that the head loss would remain the same for the two cases and was for illustrative purposes only. However, if only 1 pump is used then 2 1/2" plumbing is sufficient but if 2 pumps are used, I would recommend 3" pipe which is why I gave the example I did.

So if you go from one pump to two pumps but now keep the plumbing size the same, the head loss will increase and the flow rate will increase but not by the amount in the first example.

Your exact increase in flow rate will depend on the operating point with a single pump and that is difficult to know for sure unless you have already done some measurements on it. Theoretical head loss determination is possible but requires very detailed knowledge on the exact plumbing configuration but even then, there are things that cannot be determined. So I usually don't rely too much on theoretical head loss calculations unless they can be calibrated by a few measurements (i.e. suction and presssure). If you can measure both on the pump, then it is much easier to predict what the flow rates will be under different pump conditions.

The best way to determine what size pump(s) you need, is to first determine what is the flow rate that you actually need. With that information and the mesurements with your current pump, we can zero in on a solution.

I would like to continue this discussion on the original post so that others can benfit from the discussion so any more questions, please use the post.

Thanks


View message

carlos31
Message subject: Re: Insufficent watefall pump size !!! What size should i use Sent at: Thu Jul 10, 2008 12:26 pm
thanks for trying to explain this to me, but I am still confused.

You wrote "A single 1 HP Whisperflo with 2.3" plumbing yields 108 GPM @ 37 ft of head so two pumps with 3" plumbing would give you 216 GPM at with the same 37 ft. of head."

Should I expect 216gpm actual output OR is this only theoretical because using two pumps will increase the dynamic head far beyond the 37ft when using one pump?

Sorry, I am not catching on quickly here!


View message

mas985
Message subject: Re: Insufficent watefall pump size !!! What size should i use Sent at: Thu Jul 10, 2008 11:12 am
Carlos,

One thing that you must remember is that dynamic head loss is dependent on flow rate. Flow rate is dependent upon the pump HP or the number of pumps you have. So the higher the HP pump that you put on a given plumbing system, the higher the flow rate and head loss will be. So a 5 HP pump is likely to have very high head loss.

Next, parallel pumps will double the flow rate for a GIVEN head loss. Pumpworld web site shows this in their tables. However as I pointed out above, dynamic head loss is dependent upon flow rates so what will happen is that double pumps will increase the flow rate but head loss will also increase. So you won't get double the flow rate but you will get much more than a single pump aand the pumpworld website shows a 40% increase.

So you can compare a double pump head curve with the 5 HP head curve and decide which way to go but usually, the double pump method is more efficient.


View message

carlos31
Message subject: Re: Insufficent watefall pump size !!! What size should i use Sent at: Wed Jul 09, 2008 3:35 pm
mas985,

I am nearing the time to pick pumps for my waterfall and in the post below you suggested two parallel pumps and figured some curves for me. I was considering the 1.5HP Whisperflows to increase the flow by another 40 gpm. A local company suggested a 5HP Challenger but its pump curve doesn't seem to match up with my 40ft of head. I will be asking them about this. However, while searching for alternative pumps I found this site.

http://www.pumpworld.com/parallel_pumping_identical.htm

This site indicates that two pumps in parallel will not double the flow as your calculations indicate, but they don't provide any rationale for their assertions. Does this web page make any sense to you?

thanks,

carlos31



Subject: Insufficent watefall pump size !!! What size should i use


mas985 wrote:
Unfortunately, most medium and high high pumps do not publish flow rates that high in the head curves so it is uncertain if they could achieve that. Theoretically, they could with low head but the pump itself adds a certain amount of head loss so even if the plumbing were zero head loss, there is still head loss in the pump and the pump would max out at some flow rate. I suspect this is the reason that most pumps don't show flow rates above 180 GPM. The 3 HP Northstar goes to 180 GPM but at 60 feet of head. So it is not clear what happens to many of these pumps above the published levels and extrapolation can be erroneous.

The Pentair waterfall pumps are very low head waterfall pumps. The Hayward Tristars have good flow at higher head values but the specs start at 40 ft of head so it is difficult to say what happens at lower head. Discussion with the pump manufacture may help but I suspect that they don't even measure the performance beyond what is published in the head curves.

Another option is to use two matched parallel pumps. You can get away with 2 lower HP pumps while doubling your flow rate for a given head loss. A detailed analysis of this type of configuration is a bit tricky but can be done. The pumps see the same head loss which is determined by their combined flow rates. One easy approximation is to assume a single pump with plumbing of 2.3" diameter or 3.6x the equivalent length and then double the resulting flow rates. So a single pump with 2.3" plumbing will have the same head loss as a parallel pump system with 3" plumbing. So for example:

[EDIT] Corrected these numbers such that the plumbing would yield 26 feet of head at 175 GPM.

A single 3/4 HP Whisperflo with 2.3" plumbing yields 94 GPM @ 30 ft of head so two pumps with 3" plumbing would give you 188 GPM at with the same 30 ft. of head.

A single 1 HP Whisperflo with 2.3" plumbing yields 108 GPM @ 37 ft of head so two pumps with 3" plumbing would give you 216 GPM at with the same 37 ft. of head.

The pumps are still running at a low head loss but it is within the operating range of the pump this way. The pumps will run quieter than a single high HP pump and with less energy use (2x1.3kw vs 3.1 kw for 3HP). The downside of course is the cost of two pumps. However, this technique could also be applied to lower cost less efficient pumps as well such as the Pentair Superflo.



BTW 3" 90's are closer to 8 feet of straight pipe.

 

[carlos31]

Maybe this is the better way to attack the problem.

If I shoot for the maximum possible flow using my plumbing maybe I can size the pump and just adjust the flow down with a feedback valve that loops water back from discharge to suction when the valve is opened???

What is the most water I should try to move through the 3 inch waterfall plumbing? (The water exits through a slot 3/4 inch high by 60 inches long cut into a 3 inch pipe.) ((Starting to get bad feeling that I should have used 4 inch))

I can increase the suction side (3 two inch pipes into a three inch pipe) by adding the two inch main drain into the piping and increasing to a four inch pipe. On the return side I can only increase the pipe to 4 inch for 130 feet of the run. The last 30 feet is encased in concrete.

thanks,