PDA

View Full Version : Low Design Flow During Fire Flow



D.Assan
February 28, 2017, 10:19 AM
I have a model whose results are puzzling. I applied a fire flow of 1,500 gpm to all Junctions along the main line and the two (2) fire hydrants. The change in elevation from pump location to the most extreme fire hydrant is +55 ft. The design flow obtained at all junctions on the main were below 1,500 gpm. I want to know what I can do to get the design flow up over 1,500 gpm. I have attached the fire flow results and an image of the model for consideration.

Vladyman
February 28, 2017, 10:22 PM
Your critical nodes are J10 and J26. It would be helpful to know which are those nodes.
Also the fireflow analysis would be helpful; to verify static pressure - residual pressure drop.

D.Assan
March 1, 2017, 10:54 AM
Thank you for the reply. Nodes J10 and J126 are the Fire Hydrants on the network. The results posted above is actually the fire flow analysis.

Patrick Moore
March 6, 2017, 10:26 AM
David,

One of the best ways to identify what parts of the system are causing excessive headloss during the fire flow conditions is to run a "manual" fireflow so that you can see results for each pipe and junction when the fire is occurring. You can do that fairly easy using the software by either manually placing a demand in the amount of the required fireflow on the junction and running a standard analysis or by using the multi-fireflow tab and simply selecting the single junction that is showing less available flow than is required. The Multi-fireflow tool will then add the fireflow demand on top of the existing demand and you can see results for every pipe and node in the system when the fire is occurring.

This is key in identifying what the issues are because what you want to do is to find what pipes in the system are causing high headloss. Generally if you look for pipelines exhibiting velocities greater than 10 fps during fireflow conditions that is a good place to start. Generally if the fireflow run predicts less available flow that is required this means you generally need to add pipeline capacity to the system as this will reduce the headloss in the pipelines and allow the system to provide the required fireflow while still maintaining the minimum pressure specified (often 20 psi in the US). When you can see what each pipeline is doing when the fireflow is added, you can identify where the "excessive" pipeline headloss is occurring and see where the system has limited pipeline capacity and where adding a new pipe or upsizing an existing pipe will have the most impact. This is somewhat an iterative process, but using a manual fireflow is generally the typical way that most engineers determine where the problem is and how best to resolve it. One you finalize your pipeline improvements using a manual type fireflow run, you can then rerun the fireflow tool to verify the improvements satisfy the required fireflow criteria.

One last note, sometimes certain nodes can have limited available fireflow due to the elevations of the junctions themselves (being too high for a region) or having a poorly defined search junction criteria such that one or more nodes in the search range are starting at a low pressure (say 25 psi or so even before the fireflow is added) such that this creates an unexpected limitation on what the system can provide. I would specifically check the elevations of junctions J126 and J10 as those are critical . It is likely that the elevation of J126 is much much higher than any other junction.

Since this is a single main pipeline system with no looping, your primary headloss will occur along the primary main that comes from the pump station. But you may have a very small pipeline on one of the short loops that may also be causing an issue for just a few junctions. But to get the entire system to meet 1500 gpm, you generally will need to upsize that primary main to be either an 8 or 10 inch size to get 1500 gpm available at all locations. An8 inch line will be right at just under 10 fps at 1500 gpm while a 10 inch will be less. When you have a long distance without looping you can run out of available head when the mainline is so close to 10 fps that you may need to upsize it to achieve your goals. That much one can say even without seeing your exact model to verify just based on experience. This is likely why nearly all results are in the 1100 gpm range and not 1500. Upsize the main to ten inch to get improved fireflow.

Hope this helps, please feel free to reply with any additional questions.

Patrick Moore

D.Assan
March 6, 2017, 03:05 PM
Thank you for your response. After checking out the pipe results, I see that the problem pipes are those that I inserted between two adjacent nodes in a double service connection. I put in a length of 1-ft together with the 8" diameter. Is there a way to insert two adjacent nodes in a double service without creating a 1ft pipe between them?

D.Assan
March 6, 2017, 03:15 PM
Thank you for your response. After checking out the pipe results, I see that the problem pipes are those that I inserted between two adjacent nodes in a double service connection. I put in a length of 1-ft together with the 8" diameter. Is there a way to insert two adjacent nodes in a double service without creating a 1ft pipe between them?

Patrick Moore
March 6, 2017, 03:46 PM
David,

You have to have at least a non zero pipe length for any model pipe, so any two nodes placed will require at lest some length of pipe. However, if those junction don't have demand, there is no reason necessarily for having two nodes even though it is a "double service connection". If there is no demand you could workaround this by putting a single junction on the main with two pipes splitting out from it. This way the main stays complete without the need for short pipes. If you had demand on the two nodes on the main you could still have two nodes just off the main, but connected by a single junction on the main itself.

If you have further questions on this issue please let us know.

Patrick

D.Assan
March 7, 2017, 07:59 AM
The actual demand or service is at the ends of the "Y". The demands placed on the "double service" are just transfer demands on the way to the ends of the "Y". I will try both scenarios with the nodes and let you know how it works

D.Assan
March 8, 2017, 03:16 PM
So I replaced the pump at the foot of the hill with one that has a static pressure of 99 psi instead of the current one with 74 psi and rerun the fireflow analysis and both hydrants recorded design flows of 1771 gpm and 1964 gpm. This just proved to me that the pressure at the connection point on the Main City Supply line is inadequate to provide fire flow to the top of the hill without some form of booster pump. Is there another option available?

Patrick Moore
March 9, 2017, 09:38 AM
David,

By boosting the static pressure you are increasing the headloss that can occur between the pump station and the fire while still maintaining 20 psi. The other option you have is essentially to upsize your main from the pump station to the point you are testing.

For example. original static pressure is 74 psi before the fire. This means you have 74-20 or 54 psi of headloss you can handle at your required flow before you have an issue. This headloss would occur in 2 places a) From the source system up to the pump station itself and b) from the pump station to the point of the fire. You usually can run a flow test before the pump station to see how going form normal demands to the fire demands would lower the suction pressure at the pump station and you can then (if you don't have the rest of the system) account for this by using a GPV between the reservoir and the pump suction. This valve would have zero headloss at zero flow but would increase to the value identified in the field when flow testing. You should generally try to eyeball the curve so that the headloss curve looks exponential rather than a strait line form zero to the maximum as the headloss will increase exponentially as the flow increases.

One you have that curve your remaining headloss represents what happens from the pump to the test point of where you are concerned. If you get too much headloss at your required fireflow then the only way to reduce the headloss is to a) increase the pump size of your fire pump or regular pumps to account for the headloss during fireflow or to increase the pipe capacity (make them larger) so that there is less headloss at higher flows. Generally it is easiest to increase the pipe capacity. For a long strait length of pipe like you have 8 inch at 1500 gpm would generally burn too much head too quickly and so you may need to upsize all or part of the line to a 10 or 12 inch line or find some way to create a loop of 8 inch lines to allow the fire flow to come from both directions of the pipe. Increasing the pump size is much more costly especially if the pump station already exists.

If you ran the fire without the booster on you still "could" have enough system head if you increase the pipe size of the main, which is something you would want to check first. You also need to make sure how much system headloss occurs before your subdivision is accounted for as well as this could be significant to your final recommendation. Ideally, you want to supply normal and fireflow supply without a pump and if you can this is generally preferred as fire pumps have very specific requirements such as being a specific fire pump and often have back-up power requirements so just make sure you are checking everything...

hope this helps a bit.

Patrick

D.Assan
March 13, 2017, 10:36 AM
I re-drew the network removing the short pipes and extra nodes and setting one node to supply to services. I increased the pipes from 8" to 10" to 12" and still did not achieve design flows over 1500 gpm. I'm attaching the results from the latest fire flow analysis.
388389390

D.Assan
March 13, 2017, 10:39 AM
please find the screen capture of the analysis

Patrick Moore
March 13, 2017, 10:45 AM
David,

Would you be able to contact us at support@innovyze.com so we can set up a WebEx and look at your model together?

If you can't contact support try running a manual fireflow by placing a demand on the fireflow node in question and look to see where the system headloss is occurring.

While increasing the pipe size is resulting in slight changes in the available fireflow, there appears to be something else that is your limitation that is difficult to tell with the information provided. A 10 or 12 inch line should eliminate the majority of the pipeline headloss so it should no longer be that. Another possibility is that your pump curve is the issue in that at higher flows the head drops or that the elevation of junction J10 may be severely limiting what head you have to "burn" in the fireflow. But it is definitely something besides the pipe size that appears to be limiting. Run a manual FF and look at what the pumps are doing and where the headloss is occurring. That is the only way to really identify what needs to be adjusted.


If you do email support please refer to this forum and direct the email to me and I can get with you.

Patrick Moore

Patrick Moore
March 14, 2017, 08:50 AM
David,

Glad we could meet and discuss this further via WebEx. For the sake of the forum I will post a few notes for others

1) Change in water supply. Since you had a flow test from the system and there are no existing pumps we set the supply reservoir at the HGL of the static pressure in the flow test. We then created a general purpose valve that had Headloss vs. flow based on the flow test hydrant curve by calculating the total headloss at several flows as the Static Pressure minus the residual pressure at that flow for several points to get a relatively smooth curve.
2) We then ran a manual FF to see what the pressure downstream of the valve would be at peak day (i.e. max day) demands plus your required FF with the manual FF of 1500 gpm at the farthest junction. The pressure downstream of the valve indicated an expected pressure of 34.5 psi. This mean we could only have 14.5 total psi headloss (~33 ft) in your system during a fire and still remain above 20 psi.
3) To get a guess at the likely pipe size needed we used the hydraulic calculator in the InfoWater-> Tools-> Hydraulic Calculator and knowing the max main length was around 600 ft with C= 120, Q = 1900 gpm (1500 ff + 400 gpm Peak day demand) we calculated the expected headloss in 8, 10, and 12 inch lines and found 8 would likely be too much (> 33 ft HL), while a 10 should be fine(around 14-15 ft hl) and 12 seemed a bit oversized at around 4 ft HL.
4) You were then going to change the pipe diameters and to verify using a FF run that your desired criteria would be met with these changes. This would also take into account the elevation change between the valve and the farthest node as this would also limit the maximum headloss available.

One thing we should have also checked if we had time was to verify what the static pressure at the farthest node would be when the discharge valve was at 34.5 psi because any positive elevation difference between the valve and the farthest node you will supply fireflow will reduce the head available for fireflow. If we could have 33 ft of headloss but had 30 feet of elevation rise at the farthest node for fireflow you would then only have a maximum of 3 ft of headloss that could occur and to still remain at 20 psi. The one good news is that you may not need to have a hydrant at the end of the 600 ft capable of doing that as local fire codes (you should verify) may only require you to provide a fire hydrant within 500 ft of each home capable of meeting the required FF. This could reduce you pipe length from 600 ft down to maybe 100-200 ft which could make a huge difference.

If the elevation change ends up being the constraint, then the pipe improvements may need to occur upstream of your new development such as upsizing an 8 inch main to a 10 in order to increase the available head at the Max day plus FF demand in your development. What you are trying to find is where the pipes are constraining you (i.e. as we saw the 8 inch supply lines really seem to be too small) and this may simply be upstream of your new development.

Patrick