Help in headloss calculation for fire hydrant pump 2

Head loss calculation for 100mm dia pipe

Head loss calculation for 150mm dia pipe.

Numbers look right, the pressure drop calculated is high because the 100 mm branch pipe is too small for the full flowrate. If this is a network with multiple users the full pump flowrate will not be going down the 100mm branch. Are there more than one 100mm branches?

ETA: Missed the layout drawing, pump discharge header is undersized at 100mm. Second fel3's suggestions.

I did the calcs my way, which is to use USCS units (we're somewhat primitive here in the USA ) and the Hazen-Williams Equation instead of Darcy-Weisbach. My results (see attached) are very similar to yours. Usually, I do quick calcs like using an interchangeable solutions program that I wrote decades ago for the HP-42S calculator, but here I did the calcs in Mathcad so that you could see them. Anyway, your head loss calcs are fine.

The problem is the layout of your system. The main line is all sticks and no loops and the main line is very long. The way to reduce system head losses it to loop the system. The first and most obvious fix is to add a short segment of 150-mm pipe just inside the truck entry to loop the main line. You might even want to add one or more additional loops inside the site. To analyze this now more complex system, I suggest you model it in EPANET (freeware from the US Environmental Protection Agency: or one of the many commercial water system modeling software programs that are available (e.g., WaterCad). Water modeling software has a learning curve, but I don't think setting up and running a model like this will be too difficult. If you get stuck, there are more than a few of us on this forum that know water modeling pretty well.

I will follow this post with a mark-up of your site plan to show possible additions to the water system.

============
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

Here are my suggested additions to the water system. A 150-mm pipe across the truck entrance (solid blue) to complete a loop of the main line is an obvious to-do. The dashed blue lines are other possible additions as needed to meet your design criteria. Alternatively, you should consider upsizing the 100-mm pipes to 150-mm.

============
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

Let's try uploading the site plan again.

============
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

fel3 Sir

The way to reduce system head losses it to loop the system.

As you told we have done it in our old plant but I don't know how looping will reduce the head loss if so how much head I can save from head loss can you please explain me a little ?

The first and most obvious fix is to add a short segment of 150-mm pipe just inside the truck entry to loop the main line.

Yes, we can add one loop at truck entry point.

You might even want to add one or more additional loops inside the site.

What will be the size of additional loop?
As you told we can go for 150mm pipe all over the plant ?

Can I go for the same pump head as vendor told us (70 head) ? Or can we increase a little bit around 90 meters ?

Here are the thoughts for reference,
1. Refer to your local Code, similar to NFPA 24 in US, for the firewater loop system requirements
2. The loop piping of the firewater Main is to equipped it with multiple water sources to ensure the water supply source and flow rate needed
3. Install sectional isolation valve on the loop to ensure to ensure the maximum number of hydrants in each section per Coder
4. Depending on the total water flow and number of hydrants needed on the incident, the water main may be designed accordingly. Per NFPA Code, the minimum water Main is typical 8 inch or larger and branch is 6”.
5. Check and confirm with your local AHJ if any additional requirements

Fire mains should always be looped. As explained by others it does help with the pressure drop but the primary reason is for reliability. It allows you to do maintenance on sections of the loop while supplying water from the other side.

I cannot see on the plot plan where the fire water storage tank and the pump are. Probably just my bad eyes. But wherever they are, it seems that you are calculating the pressure drop through the 150 pipe using its full length. You should only use the length from the pump to the junction with the 100 pipe.

It does seem that the 100 branch pipes are undersized but once you recalculate with the loops in place and with the branches fed from both ends the situation will improve markedly. Upgrading the branches to 150 NB would go a long way to solving your problem.

I see in your calculations you have entered the pipe diameters as exactly 100 and 150 mm. You should check what the effect of using the true inside diameter is. In turbulent flow the pressure drop varies inversely with the 5th power of the diameter so a Sch40 100NB pipe with an actual ID of 102.3 mm will decrease the frictional pressure drop by 10%.

Katmar Software - AioFlo Pipe Hydraulics

"An undefined problem has an infinite number of solutions"

katmar sir

Thanks you so much sir for your information.

I will check my pipe diameter. I did seperate calculation for 150mm and 100mm.
Anyhow it reached 5 bar head loss so I will have to increase my pump head.

Asisraja…

Looping reduces head losses because it splits the flow among several pipes instead of requiring all of the flow to move through a single pipe. The effect is huge and you can easily check it yourself. Take either one of your calculations, cut the flow in half, and look at the resulting head loss. This is equivalent to having two identical parallel pipes delivering the full flow from Point A to Point B, which is a simple loop. Per Darcy-Weisbach, the head loss will now be 1/4 what it was before. Per Hazen-Williams, will be about 1/3.5 as much. The difference in the reductions is due to the exponents used in the two equations. In the Darcy-Weisbach equation, head loss is proportional to the square of the velocity and thus to the square of the flow rate. In the Hazen-Williams Equation, head loss is proportional to the velocity (and to the flow rate) to the power of 1/0.54=1.852.

However, the actual reduction in head losses in a typical looped system will not be as simple as being 1/4 (or 1/3.5) of what it was before because you will rarely encounter the symmetry that I used in my example above. If you model this in software, you need to check fire flows at one or more "hydraulically remote" locations. If your site had significant vertical relief (which is obviously not the case here), you would also need to check fire flows at one or more high elevation locations. Basically, you set a fire flow at one critical fire hydrant and see if the system meets your hydraulic and pressure criteria, move the fire flow to another hydrant and see if that works, and repeat as needed.

============
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

fel3 sir

Per Darcy-Weisbach, the head loss will now be 1/4 what it was before. Per Hazen-Williams, will be about 1/3.5 as much

Does it mean If my overall head loss is around 5 bar then my head loss is 1/4 of it is 1.25 bar ? Is this right or am I getting wrong?

Can you please explain me a little bit with my calculations?

This concept is new for me because I know only Darcy weisbach equations but not totally how it is involved in it.
Thank you so much sir.

Asisraja....

If you cut the flow in half per my example, then the head loss will decrease as I stated. However, once you loop a system, you can't guarantee that the flows will divide equally (see below). This means that your single pipe calculations will be outdated and you need to switch to water modeling software to solve the problem. There are techniques to solve simple loop systems by hand (e.g., the Hardy Cross Method), but I wouldn't force that chore on even my worst enemy.

If you have a completely symmetrical single loop and place a single demand exactly opposite from the source, then the flows will split equally. But, if a single demand is somewhere else on the loop, the flows will split so that more water passes through the "shorter side" of the loop than through the "longer side", so as to balance the head losses on each side of the loop. With multiple loops and/or multiple demands and/or multiple sources, etc., it gets to be far too complicated to sort out by hand, which is why I recommend using water modeling software.

There is nothing wrong with using Darcy-Weisbach (EPANET can use it as well), but the most commonly used equation for evaluating water system (at least here in the United Stated) is Hazen-Williams.

============
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

I've always worked on the basis that the worst case is a demand half way around the loop.

Anything else results in a similar but lower pressure drop.

To check just do one at 25% around the loop assuming no flow from the other half or maybe 10% flow.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

LittleInch…

For a single loop, your strategy is appropriate and it is what I do for a "quick-and-dirty" analysis. Here is a detailed description I wrote about a decade ago about how to do this:

If Asisraja adds the extra pipes that I suggested, then modeling this on the computer is the better approach because he really needs to know how the system would function.



============
"Is it the only lesson of history that mankind is unteachable?"
--Winston S. Churchill

It seems that Hazen-Williams is fairly universally used for fire water calculations. Not because it is better than Darcy-Weisbach but because the insurance companies have been using it for 100 years and they trust it. As a process engineer, I have more trust in Darcy-Weisbach. So I have used D-W for my designs and then confirmed the results with H-W. This way I know it will work and keep the insurers happy. Check with your local fire chief and with your insurers which methods they will accept.

For the simple systems that I have worked with I have taken LittleInch's approach and made sure that the worst case is satisfied. Even with a looped system the pressure drop through each side of the loop will be the same (since they share start and end points) and it is easy to calculate the flow through each side to get the total flow.

But to balance a slightly complex system by hand can take a whole afternoon and just when you are finished someone will ask "what will happen if we add another hydrant?". So I have to agree with fel3 and recommend that you use a proper modeling package.

Katmar Software - AioFlo Pipe Hydraulics

"An undefined problem has an infinite number of solutions"

fel3 Sir
Thank you so much for your valuable information. I am not good at modeling software I know only basic AutoCAD anyhow I will try modeling software as you told here. I have to know how head loss plays a major role here.

katmar sir
I have more time here to work with because still my management doesn't conclude the final design so if I design properly it will be really helpful for me to understand more core knowledge in this system.

Thank you sir.

Hi,
I don't think the safety bureau or the fire brigade, or the Insurance company will accept your design. You need to have a loop to mitigate the risks in case or leak or maintenance of the system. This is key. You need to review your system with more isolation valves, more piping.
My advice, discuss with the authorities and get the approval on design.
Note: Insurance company can provide guidelines to calculate network and/or can assign engineers to support you.
Based on 10 years' experience in China.
Pierre