Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in
  • Congratulations IDS on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Fire pump question 1

Status
Not open for further replies.
fpst

fpst

Computer
Jan 20, 2012
109
Ok I have a pretty unorthodox question here, I'm not considering doing this for any job it's just for the sake of learning.

Fire pump configuration (something along the lines of this):

49FUQya.png


Ordinarily you would use a pump based on the full system flow demand given that any hose demand is taken care of before the pump, but perhaps a pump of that size is too much for other reasons (suction requirements for example). It would be nice if you could use a pump on only a portion of the water feeding the system, boosting the pressure on only a portion of the total water feeding the system, and having this boosted water merge with the city supply water and average out, being enough to satisfy the hydraulic requirements, without using a full size pump.

I haven't seen this done so I assume there's a reason it can't be done, just looking for the reason(s).
 
Replies continue below

Recommended for you

Sort by date Sort by votes
I don't think what you are diagramming makes much sense. But, I have done similar where I take a portion of the system from the city supply, and then supply other areas from the pump discharge.

1 - High rise building where pressure reducing valves are required: If the city supply can handle the lower few floors without the fire pump, I will design them off the city pressure and save the pressure reducing valves. This just depends on how much the "duplication" of some bulk piping will cost regard to the number of pressure reducing valves you can save. You always have to look at both sides of the equation.
2 - Existing systems where a pump is installed and something is added. For example, I recall one job where there was an existing 1000 gpm pump and there was a change that required in-rack sprinklers. The new demand of overhead + in-rack exceeded the maximum capacity of the pump. However, I was able to take the in-racks off prior to the pump and then use the pump for just the overhead system. I just calc'd everything so that the rack and overhead was balanced at the point of connection where they combined and sent it through to the city supply. This was interesting because the others involved in the project were looking at upsizing the incoming fire line and upsizing the existing pump. Sometimes, thinking slightly outside of the box can bring solutions.

In your situation, you have to realize that pressures will balance to the higher pressure at a flowing junction. I really can't see how any of what you are describing has any real or practical application. Maybe some that are much smarter than I can figure out how your diagram may be a good idea.

Travis Mack
MFP Design, LLC
"Follow" us at
 
Upvote 0 Downvote
I'll admit that I'm still very much unlearned regarding fluid dynamics outside of how it applies to a typical sprinkler system every day in order to get it to satisfy a safety factor in software. The chances are high that any situation I propose has erroneous beliefs that it's standing on - therefore not applicable in the first place, so please keep that in mind when trying to rationalize what I'm diagramming.

Let's say that using a 6" fire pump at 150% causes your city suction pressure to fall below what the water utility will accept.
Let's also say that using a 4" fire pump will not give you the required flow at 150% of the pump capacity required for your hydraulically remote area.
The normal line of thinking, I suspect, is to suggest a tank at this point as an alternative source of water supply or add pumps in parallel. I could be wrong as I haven't really gotten into sizing/deciding on tanks and fire pumps in detail yet.

However, what my diagram attempts to do is divide the single city supply into two water source paths - one path which passes through a 4" fire pump and one path which does not. What I would logically expect to happen, not having intricate knowledge of fluid dynamics, is for the city supply to split into proportional volumes at the 6x6x4 tee upstream the fire pump, the 4" pipe path having less volume due to smaller pipe size than the 6" pipe path. The proportional flows in each path would travel until they met at the downstream 6x6x4 tee junction, then the 6" pipe path and the 4" pipe path would equalize pressure - the entire city supply pressure effectively being boosted, due to the 4" pipe path, without it all having to travel through the 4" fire pump. It would not be as good as it all flowing through a 6" fire pump, but in the case that all of it flowing through a 6" fire pump violates the water utility suction rules, it would be an acceptable workaround. That is the idea or theory I had, I am 99% sure it is either folly in the first place or won't work for some reason, but I thought I might as well ask.
 
Upvote 0 Downvote
I like your questions. It is great to foster the learning environment. I understand now what you are trying to get, but I still can't see how it would be in a real world situation, or how you would go about guaranteeing the proper flow through each leg. However, I would be happy to be wrong on that statement.

One thing to consider when you get into tanks is to use an automatic fill valve on your tank to help reduce the quantity of stored water. Let's say you have a 500 gpm demand (cut up dry system caused stupid high demand) for 30 minutes. Your city supply can only provide 350 gpm at 20 psi (real world example with this). Pump and tank required. Now, you can do 15000 gallons of storage. Or you can use 350 gpm from the city dumping into a 5000 gallon tank. In a 30 minute duration, the city supply will provide about 10000 gallons and the stored water makes up the 5000 gallons. So, you can save a bit on the cost of the tank.

Travis Mack
MFP Design, LLC
"Follow" us at
 
Upvote 0 Downvote
Well, if it were allowed and if there was a mathematical way to determine how much flows through each path, you could in theory use a pump that is just big enough to give you the pressure you need for your hydraulically remote area - and no more - in conjunction with a pipe size, such as 4", 6", 8" etc.

This all came up because my shop had a sprinkler design that was really close to being possible without a pump, but just needed it for about 5-10 extra psi. A very beefy non-pump system might have worked, but it would be outrageously costly. Forcing the entire supply flow through a tiny pump probably isn't going to work and isn't a good idea, and a large pump is expensive not just in itself but also the electrical required for it etc. Well, the idea was if you could use a tiny pump in a piping loop (like the diagram in the original post) to marginally boost the entire supply without the entire supply going through the pump, you get the best of 3 worlds - 1. a bid price that doesn't make you lose the job because you used a tiny pump and everyone else used a large one, 2. you get the fire pump inspection income out of it by installing a fire pump rather than using a very beefy system by having the pump vs no pump, and 3. your sprinkler system remains as simple and hydraulically streamlined as it can be giving the best value to the customer plus making it easier on the little designer everyone forgets about :)
 
Upvote 0 Downvote
"Forcing the entire supply flow through a tiny pump probably isn't going to work and isn't a good idea"

I don't understand this. If the system demand is 300 gallons, I could use a 250 gpm pump, instead of a typical 500 gpm pump. This leads to smaller valves and equipment in the pump room which helps shave costs. Why would you not use the smallest pump that meets NFPA 20 criteria? Maybe I am not understanding your full statement.

I use a fair number of 300 gpm vertical in-line pumps for residential/mercantile buildings. The QR reductions often permitted in these types of projects can lead to much smaller pumps than were traditionally used in the past.

Travis Mack
MFP Design, LLC
"Follow" us at
 
Upvote 0 Downvote
What I mean is, for example, lets say your system demand is 300 gallons, you could use a 250 gpm pump, but it boosts the pressure of the system by 30 psi, and you only need 5 psi more - which could be accomplished by lets say a 50 gpm pump that boosts the pressure of only a portion of the supply (like the diagram)

(plus in our case, the minimum size pump we needed was dangerously close to the minimum suction allowed by the water utility, but that isn't the main reason here).

lets also say that at some point, x gpm pump requires much more cost in terms of electrical power to the pump room, if they even have it in that area. this is another reason that using an even smaller than typical pump would be nice.

what I mean by "tiny" pump, is a pump that is smaller than the minimum size pump that handles the entire supply of a sprinkler system. If your minimum size pump is 250 gpm for a 300 gpm sprinkler demand, that is the minimum size pump, a tiny size pump to me would be something like a 200 gpm pump, or 150, 100, 50, etc.

I threw a lot of pump sizes around not knowing if they exist in that size or not or if they are practical in those sizes, so apologies if that confuses everyone.
 
Upvote 0 Downvote
Not an engineer but you are missing some check valves ?

In the diagram if the pump kicks on , it would pressurize the city main?

If there was a check valve not allowing the fire pump pressure into the city water main, than whichever system had the higher pressure would keep the other check from opening, giving you only pressure from one side ?
 
Upvote 0 Downvote
you could simply put a check valve before the first 6x6x4 tee junction to prevent pressurizing the city main, if there's not one already
 
Upvote 0 Downvote
Yeah. There are some check valves missing and such. Luke is in the very early stages of his career. He is trying to figure out the math and physics of some of these things.

Basically, a system like this could never work. It would be like trying to use a jockey pump to "supply" a sprinkler system. You can't only boost a small section of a system and try to get that to boost the entire system. This is basically a system of two supplies in parallel.

If you are in parallel, the flows are additive, but the pressures are not. But, Luke is trying to see if he can apply an idea to fix a situation. In some cases, you will come up with a brilliant idea. In others, you fail miserably. This example is one of the latter.

Nita, we have so many that blindly follow what goes in and out of the computer software without any understanding of it. It is good to see a very green sprinkler tech trying to think outside the box. If he ever put this loop in his calculation program, he would see it will not work. The point though is that you need fresh thinking to find fresh solutions at times.

In the example above where I stated to take the rack supply upstream of the pump, the salesman/project manager and local AHJ did not think it was valid simply because they never saw it before. We all sat down, went through the calculations. eventually they came to see it was very similar to taking a hose allowance upstream of a pump. All we were doing was taking off water upstream, but still made sure the city supply had enough volume for all to work.

The fresh thinking concept is often lost in our rapid paced culture because we often don't have time for trial and error. Heck, we barely have time to do things once, let alone two or three times.

Travis Mack
MFP Design, LLC
"Follow" us at
 
Upvote 0 Downvote
I think the main question is, before we even find out if this is allowed per any standard or code or valve configuration, is how fluids behave in a check-valve loop when one leg of the loop is being pressurized by a pump.

normally what happens in a loop where the check valve is there but not the pump (just a typical loop we use all the time) is that the friction loss of the varying pipe sizes (6" and 4") and the pressure loss of the check valve itself causes loss in the residual pressure of the overall flow, when it joins back at the second tee junction.

In my mind, it seems perfectly logical that the fire pump on one leg of this loop is simply negating the pressure loss due to the pipe wall friction and check valve on one leg, and then some, so that the residual pressure is higher on this leg, and then they meet at the second tee junction and average out (equalize), just as they would if the fire pump wasn't there.

however, there could very well be some factor I'm missing, and none of the text books seem to address loops with check valves in them (I'm not entirely sure, from a fluid dynamics standpoint, that the check valve is even necessary for the effect to happen, even though it would obviously be required in a real install).
 
Upvote 0 Downvote
Have Art make you do a hand calculation of a loop. The properties of fluid dynamics state the the friction losses in both legs of the loop must be equal. That is because you can't have two different pressures at a junction. The pressures will balance.

If you have access to The Fire Protection Handbook, look at the section on hydraulics. It does a great job of explaining how to calculate flows through a loop. You will learn how to do equivalent lengths of the pipes so that you can get the pressures at the loop outlet to balance.

I think it is just a property of fluid dynamics that you don't have grasped yet. Art can get you there.

Keep up the effort. If you need more help, we can set up a "goto meeting" and I can help walk you through it. This was a decent chunk of material we did in class last quarter about this.

Travis Mack
MFP Design, LLC
"Follow" us at
 
Upvote 0 Downvote
It appears to be the case that each leg of a loop which has check valves in each leg of the loop must have identical momentum at the discharge-side tee in order to not close a check valve on one leg

This makes some sense to me intuitively - but there is something peculiar going on. If the above is true, does that mean you can only have pumps which have identical characteristics in parallel? Can you even have pumps in parallel (intended to operate simultaneously) at all? If you can, how can you ensure that one pump doesn't perform differently than the other in reality (due to manufacturing imperfections) causing one to create more pressure than the other, and close the check valve in the other leg of a parallel loop?
 
Upvote 0 Downvote
The set up shown is what you have on every booster pump taking suction from a city water main where you have a bypass.

You've drawn these up before but with a check valve on the bypass.

with the pump running there will be a very slight increase in pressure at the pump discharge but it will be extremely small. Actually the boost will be equal to the head loss developed from the discharge of the fire pump to the pump suction flange.

If the pump can develop 100 psi boost and the total head loss developed was 10 psi we would expect to see the discharge side of the pump to have a 10 psi higher pressure than the pump suction flange.
 
Upvote 0 Downvote
This is what I've got so far

all pumps being identical,
two pumps in series increases water pressure beyond one pump
two pumps in parallel increases water capacity beyond one pump
two pumps in series and two pumps in parallel all tied in together, increases water pressure and water capacity beyond one pump
one pump with a bypass, the bypass having a check valve, will keep the bypass-check-valve closed (no flow movement there)
one pump with a bypass, the bypass not having a check valve, some of the water will flow in a circular path through the pump over and over again, and the overall pressure gain is almost neglible, as if the pump wasn't even operating

why and how all of this is true I still don't understand. I can accept in principle they are true, but it nags me not knowing why a check valve will close if a single pump is applying pressure to it, but will not close if two pumps in parallel are applying pressure to it on both sides - differing pressures because of differently sized pumps.
the answer, logically, has to be that the pumps in parallel somehow balance the pressures, but then it nags me how pumps can achieve this "balancing pressure" effect, but passive loop + pump loop legs cant.
 
Upvote 0 Downvote
Remember pumps follow a flow curve. As more water flows through a pump, the less pressure you have. So in your example above, even with check valves - initially, water will be flowing through the pump leg because the pressure will be higher, closing the check valve on the 6" leg. As the pressure gets less with more volume, you will eventually have the pressure drop to equal the pressure in the 6" leg and that check will open. That will then allow the pressures to balance.

Your primary assumption was that the pump leg would not be able to supply the total demand. So eventually the pressure in that leg will drop so the 6" check valve can open.

This is no different than a jockey pump and main pump set up. With a leak or maybe one sprinkler flowing, the jockey pump kicks on. It is pumping at a slight higher pressure than the fire pump. Eventually the jockey can't keep up and pressure drops downstream. The fire pump then kicks on and shuts the check on the jockey pump leg. Identical to your setup to start this.


Now,pumps in parallel are typically designed with identical flow characteristics. For example, I am working on a job that requires 6000 gpm capacity. We have two 3000 gpm pumps in parallel. The pressures are identical. There is a third 3000 gallon pump set as a backup. The controllers are set so that if pressure drops beyond the jockey capabilities, the primary pump comes on. As that can't keep up, the secondary pump comes on to add more flow capacity to the system. The third is a backup in case one of the first two fails or is out of service.



Travis Mack
MFP Design, LLC
"Follow" us at
 
Upvote 0 Downvote
Well, basically isn't this a normal fire pump set up ?

Fire pump with by pass line?

Yes pipe sizes are different and trying to use city water pressure to design.

A reason it has not been done , besides the engineering problems, is how to calc it ?
 
Upvote 0 Downvote
Travis "Have Art make you do a hand calculation of a loop."

Whenever I drag out my N^1.85 paper along with my trusty $20.00 pocket calculator with the old worksheet Luke's eyes glaze over and I can tell he's thinking "I wish I was anywhere else but here, this is so 1970's like doesn't this old fossil know about computers today?"

I guess I gotta face up to the fact if you're under 35 you never experienced a world without computers.

We've talked about it here before but I started doing calculations in 1976 the old fashioned way because there was no other way. Over time I got good at it; trees were easy, loops were easy and grids were easy to but you had to know how to set them up. I always used a ladder grid, not the best performers but they were a whole lot easier than center side feed grids which would drive you insane. Last one I did was 1980 and it took me two days.

I show look this Flow Test and Problem with The Solution and Luke just tunes out. He'll go into his office, fire up AutoSprink, draw something up that will do the solutions for him and it's done.

I am glad I learned how to do it the old fashioned way because I believe it gives me a greater understanding of what is happening but I must admit I learned the old fashioned way because there wasn't any other way back then. If I had had HASS in 1976 I would never have learned the old fashioned way.

Today there's a growing number of younger layout technicians that don't have experience doing hand calcs and to me they're missing something and if I had my way they'd all be doing hand calculations the first year but that isn't feasible in real life.

Yesterday we were discussing what it would take to put sprinklers in the apartment attic and I smiled when Luke wanted to know how I would know if something would work or or not.l I think he wanted a formula or something similar but I didn't have one. End head pressure, elevation and head loss through a backflow and I got it pretty close... no calculations just past experience.

But fresh minds are good too. Sometimes we get stuck in the old ways of doing something and the new guy will approach a problem from a different direction that you never thought of before... I gotta give credit where credit is due and Luke has done that and he's won some as well.

But this thread does it... he's going to learn how to do full hand calcs using nothing but a pad of paper, an N^1.85 graph and $20 pocket calculator. Any other old timers out there that agree with me?
 
Upvote 0 Downvote
I don't want to say I am an old timer but I agree. In my class last quarter we had to do hand loops and hand tree. We were allowed to program an excel spreadsheet or just so by paper and pen.

You learn balancing better when u do hand. A recent foam job requires crazy balancing. I followed a similar plan as I would have done by hand. This time it took me two hours to get the system to work. I did one like this last year and it took two days. So, following the thought process of hand calcs can be vital.

Have him calc a simple tree. 4 heads on one side. 2 on the other. It will teach equivalent k factors and balancing. Then have him calc a simple loop with same size piping. Lastly, a simple loop with say a 4" leg and a 6" leg with different lengths. That was basically a few questions on the final exam I completed a month or so again.

Travis Mack
MFP Design, LLC
"Follow" us at
 
Upvote 0 Downvote
Status
Not open for further replies.

Similar threads

jjmoore73
  • Locked
  • Question
Fire Pump Sizing 1
Replies
5
Views
506
UFT12
UFT12
Jeng84
  • Locked
  • Question
Fire pump potable water requirements
Replies
3
Views
374
Heaviside1925
Heaviside1925
Erfan z
  • Locked
  • Question
Fire Pump Motor
Replies
5
Views
223
LCREP
LCREP
jjmoore73
  • Locked
  • Question
Fire Pump Maximum Static Suction Pressure
Replies
5
Views
405
jjmoore73
jjmoore73
Martin Senen
  • Locked
  • Question
Fire Pump Test using spreadsheet
Replies
0
Views
222
Martin Senen
Martin Senen

Part and Inventory Search

Sponsor

Back
Top Bottom