Designing water reticulation systems

[Struggling]

Hi,

Im currently doing an assignment where i have to design a water reticulating system for a small suburb.

I have planned out all my distribution mains and my reticulation mains and services

I also have the water demand(Q) for the houses using the Australian Standards for plumbing and drainage.

So i know my Q value from Australian standards given by the formula:

Q=3.637*10^-5 H^0.555 D^2.667

where:
Q=Flow rate (L/s)
D=pipe diameter
H=head loss gradient in meters per 100m given by: (h*100)/(L*1.5) where h= Head loss in meters head, L= index length in meters and 1.5 is a factor allowing additional head loss through fittings

So for example (this is the calcs i did on one of the reticulation mains with a dead end serving 3 houses)
If i have
Q=1.03 L/s
L=50 M
h=1.1

1.03=3.637*10^-5 * 1.46^0.555 * D^2.667
D = 43.16 mm

now using this with V=Q/A i get V to be V=0.709 m/s

Am i doing this right? because the velocity seems to be very slow(even though the maximum allowable is 3m/s anyway)

and am i going about it the right way? i thought start from the demands from the houses and work backwards to the distribution mains that way i can calculate all the pressures,velocities and flows needed at the entrances of the reticulation mains.

and if this is right, what formula do i use to work out the pressure needed to at the entrance of this pipe to supply this flow and velocity? I've been reading fluid mechanics books all week and i can't seem to crack it.

Thanks in advance

 

[FredGarvin]

You seem to be on the right track. I didn't check all of your units though. Just by gut feel that velocity doesn't seem too far off.

As far as the pressure needed, I would do as you are. You started at the demand and calculated back all of the losses due to fittings, line lengths, etc... You know what pressure you want at the end for the users, now add on the pressure losses due to your piping. That will give you the minimum required at the supply to produce that flow. You can look to a modified version of the Bernoulli equation for help:

Look here under pipe flow calcs:
http://www.roymech.co.uk/Related/Fluids/Fluids_Pipe.html

Also check here:
http://odlpc.oum.edu.my/v2/tutorkits/EBMF4103%20Fluids%20of%20Mechanics%20for%20Mechanical%20Engineering/EBMF4103%20(Chapter%208)%20Fluid%20Mechanics%20for%20Mechanical%20Engineering.ppt

 

[James3849]

Hi,

Sounds like you are on the right track. I'm actually a graduate mechanical engineer designing water reticulation for a mine site at the moment and can tell you from experience that those sort of velocities are correct. Usually you would want to limit velocities to somewhere between 0.7m/s and 1m/s, this is done to limit head loss due to pipe friction and fittings, because as you no doubt know, head loss is proportional to velocity squared. Also keeping fluid velocity low is beneficial in terms of mitigatin any detrimental effects that may result from water hammer in the lines.

On another note, you should probably do your final calculations using standard pipe sizes. The next suitable size to the diameter you indicated is 50mm nominal diameter.

Hope this helps:

P.S. I am interested as to which Australian Standard you found the equation you noted

Q=3.637*10^-5 H^0.555 D^2.667

 

[James3849]

Sorry,

Missed the last part of your query. For pressure requirements, you should be investigating the most hyrdaulically disadvantage building in your network, in this case, the building farthest away. Usually city councils reccomend that around 200kPa - 220kPa is more than suitable to supply water amenties at a residential household, so getting this pressure to your most hydraulically disadvantaged facility is your goal after taking into account the pressure losses involved in getting the water to its location.

 

[Struggling]

P.S. I am interested as to which Australian Standard you found the equation you noted

Q=3.637*10^-5 H^0.555 D^2.667

Its from the AS 3500 Plumbing & Drainage Standards, the reason I am sort of worried about that formula is it says its for rapid sizing of pipes for residential areas.


i also just realized i can't use that formula because
H = h*100/L*1.5

and h is head loss which i can't work out without V or D

 

[James3849]

There are no equations that will take you step by step through the design of a reticulation system because obviously water demand varies. The equations given in the standard are just guides for you to get your head around what sort of numbers you should be getting.

My suggestion is to make a spreadsheet. If you are only supply 3 households, you should design your piping to accommodate the maximum expected demand for the houses plus a little bit extra (imagine all showers, taps are running and toilets flushing). AS 3500 will give you typical flow rates for showers and the like, but also gives you probable simultaneous demand for multiple dwellings in table 3.2. (3 dwellings = 0.88L/s). Therefore you should design your piping for 1.5L/s for example.

With this flowrate you can select an appropriate pipe diameter that will limit your fluid velocity. With this velocity you can then calculate head loss due to friction and fittings, and if these losses are too great, tweek your pipe size a little until it is okay. Don't forget to use standard pipe sizes and take into account any change in elevation too (especially if the fluid has to travel uphill).

Hope this helps

 

[Struggling]

yes it does help. I sort of figured out how to do it, made a spread sheet but then realized i had estimated my demand wrong. We have estimated each household uses 42.5 L/hr during the peak hour(8-9am). now the new problem i have is working out the flow rate in m^3/s because 42.5 L/hr is 0.0000118 m^3/s and iam convinced it isn't as easy as converting 42.5 to m^3/s(and because the pipes turn out to be absolutley tiny) but if i convert that to l/s i get 0.0118 l/s and my pipe sizes turn out pretty decent for 20 houses in 1 street with pressure losses turn out to be DN 350 as opposed to DN 600 we were getting earlier.

because in Q=VA Q is m^3/s correct? or can i use l/s for Q?

 

[James3849]

I'm not sure how you calculated your demand as I would be going with the what AS 3500 recommends for multiple dwellings. To answer your question, you have to use m^3/s, as velocity is m/s and area is m^2.

DN350 and DN600 are very large pipes and you wouldn't see those in residential applications.

You must size your pipe by figuring out pressure loss. It is pretty much a guess and check method. Keep changing the internal diamter, calculate the velocity and then figure out head loss. If head loss is large such that amenities won't function properly, increase pipe size to reduce head loss.

You cannot size your pipes on Q = VA using Q has your calculated peak demand per house, because maximum Q is pre-determined by your pumps at your pump station. They are designed to spit out a lot more than 0.0118L/s x 20 houses.

You should also conceptualize that there is alway pressurized water in reticulation piping, and when a tap is turned on, the pump station will kick in again and return the system to its designed mains pressure.

I would be surrprised if you mains piping would be larger than DN150. Branches to dwellings may be smaller.

 

[Lindy@Amida]

I am currently working in partnership with another agency to recruit 3 engineers with a BSc/ Civil Degree or a PR Engineer with enough site experience in water reticulation for their client. The positions are based in Cape Town, Johannesburg and Durban.

Do you know any engineers that would like to be considered for any of these positions?

Please do not hesitate to contact me on: lindy.miller@amida-recruit.com or tel: 021 8546511