Pressure Booster Pump Calculation 1

If you are happy with 21 psig outlet pressure and your outlet is lower than the pump centerline, yes, and if you have zero pressure at the pump inlet.

Independent events are seldomly independent.

Not quite sure what "dynamic" losses are, but essentially yes. However your system is in fact quite complex, single feed but 11 offtakes all with different head effects and different friction losses. You also have flow rates going from zero to nearly 60gpm. I can only assume your handbook allows for not all units being active at any one time.

You need to think how you intend to start and stop this pump and the effect it has on the apartment on the lowest level but with low friction losses. As the pump starts they could get a much higher pressure than before which could be very bad. You could think instead about a central large diam header which you keep at constant pressure so that the pump starts and stops on a pressure setting and have pressure limiting valves to each apartment so they all get the same pressure regardless if the pumps is running or not. I would also think you're much better off with a set of pumps rather than one big pump. Most booster systems of this type use some sort of accumulator so that a small flow doesn't cause the pump to start and stop many times a minute. For 11 apartments this might be quite large so allow enough space on the roof or in the service shaft for it.

Whatever you do it will probably need some adjustment depending on actual usage so allow for other pumps and fit the largest central pipe you can before the branch to each apartment, my guess would be 6" to 8".

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

Booster systems or "constant pressure" systems are very, very common packages offered by HVAC pump manufacturers. Go look at Systecon, Tigerflow, Synchroflo, Grundfos, etc....

These are sets of 3-4 pumps usually with one of them being somewhat smaller and used as a jockey pump. They are controlled by PLC logic and VFD's to keep the pressure constant at all flows.

You cannot assume that everyone is going to turn all of the faucets on or off at the same time; and while the ASPE are pretty good, you must be able to handle all situations encountered: provide constant pressure regardless of the flow. You can't do this with one centrifugal pump.

Perhaps you have already considered it, but normally there is a "simultaneity" coefficient that you have to apply on your base flow. here is a well done KSB document for preparing specs for a booster, it is in French but try to look for an English version somewhere in the web (not sure it exists however) if you are interested.

The formula of simultaneity depends on the number of apparatus used in the building. Something that I've used myself in the past and which is given on page 4/13 of the link hereby :

number of apparatus per apartment = 8 (as listed)
Total = 8 x 11 flats = 88 apparatus
Y = 0.8 / (88 - 1)^0.5 = 0.09
Base Flowrate for building = approx. 12 l/s
Calculated flow = 12 x 0.09 = 1.03 l/s
It seems to me that you are overestimating the flow.
I don't say mine is correct but at least a reason for double check.

Thank you all for the kind replies

BigInch: I think 21 psig is fairly reasonable at a faucet or a bathtub right? As for the pump inlet, the pressure won't actually be zero as the water level is higher than the inlet, but I want to design for the case where the water level drops. Also the pressure caused by the water level (static head) will work as safety factor as well.

LittleInch: Dynamic Losses are losses caused by the fittings in the network according to my terminology. I'm guessing you call them something else? I am using a variable speed pump that varies flow to maintain a constant pressure in the network regardless of the flow demanded. The pump also comes coupled with a pressure vessel to feed the system with small amounts of water so that the pump doesn't have to start whenever a single tap is opened. I'm not sure about the pressure valves, are they necessary? Won't the pump variable flow be enough to regulate the pressure in the system? I'd understand the need for them in a high rise tower. Regardless, I'm not that experienced yet but I think I'll do some calculations and get some pressures at the lowest level outlets.

DumbMac: I'm not using one centrifugal pump. As I mentioned, I'm using a pair of variable speed pumps. The demand is split 50-50 between them.

rotaryworld: I thought the ASPE tables take into account "simultaneity". My understanding is that what I am getting is the most probable flow demanded. Not the flow resulting from all the outlets flowing simultaneously. The tables already have a sort of diversity factor built it. Is this not the case? Thank you for the link. I will look into it.

Losses are all friction losses in one form or another, whether distributed in the form of pipe friction or "concentrated" in the form of pressure drop accros a valve or other fitting.

Glad you're designing it that way - should work well, but as we don't know the height or layout of your apartment block it's difficult to say what the highest pressure will be compared to the lowest, but you always need to check the extremes of the system (lowest connection) at no flow as this will be the worst condition without any friction losses. I'm with the others here that 58 GPM seems quite high and you may want to think about making your pumps more of a 70/30 split or putting in three 33% units instead of two as most of the time you will be in low flow conditions and don't really want your pumps starting and stopping fequently or operating at minimum speed for long duration.

Glad we've been of some help to you.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

LittleInch: Thank you for the response. You are right regarding dynamic losses. I guess they were given their own term because they are calculated differently than other non-concentrated friction loss.

The height of the pump above the lowest outlet is ~33 ft. So there will be an additional 14 psi of static pressure at the lowest outlet
I'm going to assume that a two taps were opened at the lowest level demanding 5 gpm.
From the pump curve: The pump pressure that is provided at a flow of 2.5 gpm is 65 psi.
Now the total pressure arriving at the outlet (ignoring friction losses) is 65+14 = 79 psi
Deducting approx 4 psi for friction losses I end up with 75 psi.

This is just a quick calculation, hopefully it is "in the right ballpark"

Why did you consider the extreme point to be the lowest outlet at no flow? Doesn't this mean that there is no pressure boosting from the pump, making it not the most extreme scenario?

Because you say you have a constant pressure pressure vessel so your pump shouldn't supply much more than your pressure vessel once the pumps pressurise the system then turn off on no flow? Then you ignore friction because there's no or very little flow. Are you keeping the fixed pressure at around 65 psig?? Looks rather high to me but if that's "normal" around your part of the workd then maybe it's what people expect.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

Thank you for the response.

Sorry I made a bad mistake up there, I took the pressure corresponding to the flow from the pump performance graph as if its a constant speed pump. It is the only graph provided in the catalog for this pump so I am guessing it shows the pump's performance at the maximum operating speed. Now I remembered that the variable speed drive will produce the required flow at whatever pressure required (within the operating limits).

My pressure at the lowest outlet should be 14 psi(static) + 21 psi (boosted main) = 35 psi. If the tap is opened this pressure should drop a little due to the friction losses incurred by the incoming flow. Now if the static pressure becomes too high, then a pressure regulating valve is required at this floor to reduce the static pressure such that the sum of the static and boosted pressure becomes something reasonable. Am I correct?

I apologize if I violated the forum policies or anything :/.
In this economy, when you get a job offer, you take it

No policies broken as far as I know.

Yes looks you are correct - I would allow space to put a presure reducing valve in and suitbale isolation if it becomes necesaary, until the whle thing starts working then you won't really know, but you've opened up something else with your comments about "will produce the required flow at whatever pressure required". If I was you I would search this forum for VFD pumps as that statement is not correct. Variable speed is good, but the pump will only do what is can at a certain speed and no more when it meets your system curve (flow resistance). Bigger flow will mean higher pressure at the pump outlet.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

"will produce the required flow at whatever pressure required".
I did not mean that it can blatantly produce any flow at any flow. I meant within certain limits. What I know is that at reduced flow the operating point will slide down the system curve, and then as the pump speed is reduced the pump curve will "shrink" on both the flow and the pressure axes till it intersects the new operating point on the system curve, if an intersection is possible.

Thanks again for the help.

quoted
I thought the ASPE tables take into account "simultaneity". My understanding is that what I am getting is the most probable flow demanded. Not the flow resulting from all the outlets flowing simultaneously. The tables already have a sort of diversity factor built it. Is this not the case? Thank you for the link. I will look into it
unquoted

EnOm,

Ok clear, but I am just surprised that the handbook you refer to (ASPE) and the KSB rules (based on NF EN 806-3) end up in a such big difference in flow. dont you agree ?

rotaryworld:
You are right. There is quite a large difference. The ASPE tables are based on Hunter's Curve, which was developed in 1924. So it's quite outdated and can overestimate water demand. But I see it in use in some of the local codes so it must have some credibility I guess. The code allows other "good engineering" methods to be used. I am still looking for some better alternative especially that now I'm about to do the pipe sizing as well. Any modernized fixture units tables that you know of?

The document you pointed me to seems very interesting but I can't yet find an English version :/

Best Regards.

From my experience, supply pressures to fixtures below about 40 to 50 psig will only assure never-ending complaints of low water pressure.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

ccfowler:
Thank you for the response.
So you say I should design for 45 ot 50 psig at the tap/shower?
What I know now is that pressure for at a standard shower or tap should be around 20 to 25 psig. Is this incorrect?

Typically, a piping system will have larger pipes from which branches go to feed the several (numerous) fixtures (faucets, toilets, ...) with the smallest branch size normally being nominal 1/2 inch. Except for some faucet and shower connections, there is usually some smaller flexible line between the individual shut-off valve and the fixture itself. If the pressure in the 1/2 inch lines at the individual shut-off valves is at least about 40 psig, then complaints will be reasonably rare. There will be some significant pressure losses through the shut-off valve and the flexible line, and this will generally get the pressure down into the pressure range absolutely required by the fixture to function properly. (Leaving out the individual shut-off valves would provide some minor pressure savings, but then the water supply would need to be cut off to EVERYTHING to be able to fix any individual fixture--complaints will escalate accordingly, so you will want to avoid this potential economy measure.)

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

Another practical suggestion:

I never specify anything but ball valves for all of the several isolating valves needed in a system such as you seem to be describing. Gate valves are perfectly good and usually less costly, but almost invariably, they get over-tightened and then become very troublesome. Cheap gate valves seem to be inclined to developing bonnet or stem leaks.

Valuable advice from a professor many years ago: First, design for graceful failure. Everything we build will eventually fail, so we must strive to avoid injuries or secondary damage when that failure occurs. Only then can practicality and economics be properly considered.

ccfowler:
Thank you for your elaborate reply.

Regards