EPANET 2 Emitter Coefficient

[Fayaz810]

Hello folks,

I just started using EPANET. So far I am doing OK but emitter coefficient is confusing the hell out of me when I simulate fire-hydrant loop.

I have a fire-flow of 3000-gpm from two farthest hydrants in a loop. I what I enter any large value, the whole loop goes at very low pressure. May be I am not entering the right coefficient value or ?

The hydrant has 2-1/2" B.S valves

 

[fel3]

Emitters in EPANET are most often used for things like fire sprinklers and irrigation heads. Although the EPANET help for emitters says they can be used for fire hydrants, I have never done so. This is because fire hydrant heads have internal parts and non-linear flow paths that create much higher head losses than would a pipe of the same ID and length--and--because there are multiple hydrant manufacturer, each with multiple models and configurations, each of which perform differently. I have lab tested fire hydrants twice for one manufacturer* and I have seen data from some other manufacturers (e.g. page 5 of this brochure:

http://www.clowvalve.com/upl/downloads/resources/product-brochures/medallion.pdf**).

On the other hand, one manufacturer I contacted a few years back refused to release hydraulic test results, which may mean their hydrants are not very efficient. With so many choices and often unknowns, it's usually easier to terminate the model ahead of the fire hydrant head and make an adjustment in the design criteria.

The normal practice in civil modeling of distribution systems is to ignore fire hydrants and laterals altogether and shoot for a sufficiently high pressure at the network node serving the fire hydrant lateral (say 30 psi) so that the outlet pressure of the flowing hydrant head must certainly exceed the 20 psi (or whatever) criteria. Better yet, is to include the hydrant laterals for at least the critical hydrants, then shoot for a lower, but still sufficiently high pressure at the base of the hydrant riser (say 25 psi). I discussed this in more detail in the following thread:

http://www.eng-tips.com/viewthread.cfm?qid=349880

(post #4, dated August 9, 2013, paragraph #8).

A few questions to help me understand your situation further, plus some comments:
[1] I assume that your 3000 gpm FF is split so that you have 1500 gpm at each hydrant. Is this correct? If not, your resulting pressures will be too low. Please note that 1500 gpm in a 6" hydrant lateral is a whopping 17 ft/s. In a 50-foot-long hydrant lateral at C=130, this amounts to 7.9 ft of head loss, not including minor losses. During a real fire, the actual flow that will be pulled from the hydrant will be almost always be significantly less that the criteria. The difference between the criteria and the real flow is a safety factor that is owned by the fire department.
[2] Are you trying to get 1500 gpm through the 1-1/2-inch nozzle? If you are, it's unrealistic. If you take a look at the pressure loss graph in the Clow brochure, you will see that the AWWA limits for flow are 1000 gpm for the 4-1/2-inch pumper nozzle and 500 gpm for the 2-1/2-inch hose nozzle. BTW, this table shows hose losses for different flows and hose diameters:

http://www.elkhartbrass.com/files/aa/downloads/catalog/catalog-e-T-09.pdf.

As you can see, 500 gpm through a 100-foot-long, 2-1/2" hose, loses 55 psi, which most water systems cannot support without pumping.
[3] If you want to model the full hydrant, then I would suggest 1000 gpm through the pumper nozzle and 250 gpm each through the two 2-1/2" nozzles (if in fact you hydrant have two).

I hope this helps.


* Same model, but for two cities with different criteria.
** I have not yet checked Clow's performance curves to see how they compare with Hazen-Williams or Darcy.

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