Water distribution for sommer houses at different height 4

[e331]

Hello

So I'm doing some initial research for water distribution system for two groups of summer houses.
We have access to water @150m in the mountain above, then there are planned 12 houses @100m and 30 houses @70m.
First I was planning to use the same tank and pipe the first 900m and then split the pipe in two.
After simulating this in Pipe flow expert I can run into the problem of possible low pressure @100m if the 30 houses are using too much water.
Do you have any general recommendations, should I just run two pips all the way or fix this issue with flow rate and pressure reducing valves?
See diagram attached
It would cost about 900m extra of 50mm piping vs 900m of 63mm and valves.
Are flow rate and pressure reducing valves prone to failures?

Thanks in advance

 

[1503-44]

Make the tank large enough to store enough water to supply maximum demand. That's what tanks are for. Fill it during minimum demand, draw from it at maximum.

One pipe will be cheaper. You don't have to install and maintain two pipelines, valves and fittings.

150m is too high for the tank. That will put too much pressure (15Bars) in the system when there is no demand. You might also use lower pressure pipe to reduce cost. Someone will turn on the tap and water will blast everywhere. Maybe two tanks will work better, one at the 150m high supply point and another tank at point 6. That can feed the control valve and pipes to houses at less pressure. Try to put the tank feeding the pipes going to houses no higher than 30m or so. You don't want more than 4 bars arriving at the farest houses.

Everything is prone to failure, worse if selected poorly for the application. Be glad you don't need a pump.

 

[HTURKAK]

Apparently 150,100, and 70 m are the elevations of reservoir and housing units. In this case, the first sketch ( showing the P3,P4,P5) is not correct..

- Use single reservoir and single PL between N5-N6
- Increase the diameter of P3 to 80 mm
- You may provide gate valves on the branch N6 for P4 and P5 and restrict the flow as necessary..

 

[1503-44]

Why restrict flow.
Gate valves will run the distribution lines at very high pressure when at no/low demand.
Control valve(s) are needed to restrict pressure to manageable levels.

How can you have 2l/s going to 30 houses and 1 l/s going to 12 houses? Consumption should be more or less equal per household. A general guideline would be 50 to 150gal/person/day consumption, which after assuming 4pp/House and probable distribution throughout the day, increasing for a maximum demand period, you should have somewhere between 3 to 6 l/s, going to each house. (Assuming 50-100gpd/pp) And that is also assuming that fire water comes from elsewhere. Does your local city, town, county, or state not have specific requirements for water supply system flow rates and specifications for all of their water distribution design requirements? Where do the no-return valves go? Those requirements are usually quite important and must not be under designed. They should also have appropriate measures for future expansion.

That would also suggest a minimum tank size of 20,000 liters.

I may not be clear on the pipe diameters, lengths and elevations. A clearer drawing would help, separately denoting each data item.

 

[e331]

First of all, thanks for looking at my situation

The design spec for my region and summer houses (no permanent residents) says:
Average demand 100 liters/person/day => ca 27 gal/person/day
Maximum dement 350 liters/person/day day = ca 93 gal/person/day
Assuming 4 persons per house and average out through 16 hours, I have a flow requirement of 1,458 liters/min or 0,024 liters/sec per house, which is quite far from your 3-6l/s
But lets assume that every house as a single tap turned on (0,1 l/s) as the max simultaneous use, then I would get 1,2 l/s for the 12 houses and 3,8 l/s for the 38 houses.
Can you perhaps point out better statistical equation for usage distribution.

In my region I have also found this design formula q_max=f_s x average usage per person per day x number of persons, where f_s for summer houses is 2,0-3,0
That would translate to q_max = 3 x 0,006 x 200 = 3,6 l/s
Is that far off from your knowledge?

Since this is a private summer house area there are no requirements of fire water.

Regarding the drawing, elevation is shown at Nodes (N5 @ 150m, N8@ 100m, N7, @70m)
Pipe sizes are shown as internal sizes 55,8mm for 63mm pipe, 44,2mm for 50mm, etc. This is shown at pipes (P3,4,5) as is the length of the pipes in meters.

Thanks again

So my plan was to design this in the following way:
1) First find the pipe size required to be able to supply the required demand 1,2 l/s and 3,8 l/s
2) Check the pressure in the system with full demand
3) Check the pressure in the system with no demand
4) Add required control valve(s) so that the pressure is not to high with low / no demand
5) Add pressure loss due to fittings and distribution system
6) Check how different usage in one are effects the pressure in the other area
7) Iterate until all requirements are fulfilled

I have added attachments with the setup and result for pipe flow expert regarding steps 1, 2 ,3
I ended up with P3 = 75mm, P4 = 63mm, and P5 = 40mm

 

[1503-44]

Just a well known fact. Average usage in the US is 75 to 180gal/pp/d
Google it and you can see 76 to 200 gal/pp/d without opening any url.
I'm OK with yor 100 gal/pp/d number, presuming you have data for your country.

The next most important question to ask is demand pattern and max flow rates.
Demand (in USA is always highest for a couple of hours in the morning and a couple of hours in the evening. For that reaason you never assume that your distribution rate will be anything close to the demand/pp / 16 hours that they are all awake. That is not at all realistic. Neither is the single tap assumption. I'd say at least 20% of the time, there could be two taps per house running. Toilet and probably a shower. Remembering that 16 hours is way off base and that the family tends not to take showers/baths from 8am 12pm 4pm and 10pm. They all do in the morning, or after returning from the swimming pool. Demand is always much more concentrated than over 16h, or 8hr. 4h maybe. That's probably average. But not 100% safe to use as your design maximum rate.
Again remember that those are statistics for large areas, based on town or city averages. The thing about average statistics is that they take large numbers of houses 1000+ for them to be accurate. You cannot apply those statistics to one small subdivision and expect to have the same averages. The smaller the number of houses you look at, the larger will be the variation in those numbers that you will have to expect.

For regions where you have no consumption data at all and you are just estimating demand, the distribution time used to find the maximum demand rate should be around 2 hours. Its the maximum demand rate that you need to design for, not the average.

Using the 100 AVERAGE gal/pp/d demand, 4 people/house and 2 hours, is 10 l/s
Using your 50 AVERAGE gal/pp/d demand, 4 people/house and 4 hours, is 5 l/s
Maybe it was no coincidence that I estimated 3-6 l/s

If you want to have a well designed robust system, almost city spec, but everyone's still happy, use 6 l/s for the main line.
Then IMO the 12 unit division should have a design rate of 1.2 * 1.33 = 1.55 l/s and
the 38 unit division should have a design rate of 3.7 x 1.33 = 5 l/s

If you want to provide a marginal system that will work OK most of the time, divide those by 2.

Now for pressure control.. thinking about that for a few minutes.

 

[HTURKAK]

I am not familiar with your design spec but Average demand 100 liters/person/day seems pretty low. Better to use Average demand 300 liters/person/day ( including garden ,irrigations etc..)
Assuming 4 persons per house ;

for 30 summer house community =

maxQd ( daily max. demand ) = 2.5 Qd =2.5*4*30*300= 90 m3/day

maxQh ( hourly max. demand )=0.17*maxQd=0.17*90=15.3 m3/hr =4.30 lt/ sec.( for design of P4 )


for 12 summer house community =

maxQd ( daily max. demand ) = 2.5 Qd =2.5*4*12*300= 36 m3/day

maxQh ( hourly max. demand )=0.17*maxQd=0.17*36 =6.12 m3/hr =1.7 lt/ sec. ( for design of P5 )

and Qmax = 1.7+4.3=6 lt /sec ( for design of P3)

Ref. DVGW

You may assume the figures as a different opinion or check with other applicable norms.

 

[1503-44]

I think I would make the pipe 75mm for both the main and branch to the 38 houses. 40 is OK to the 12 houses, but I'd make it 63, just for an extra bit of mechanical strength. I'm sure you can afford it. One tree starts growing in the wrong place and ... Plastic pipe has high temperature expansion and contraction coefficients, 10x steel. A few winter-summer thermal cycles and you can wind up with high tension. Something I don't like about 1000m of 40mm. I have trouble with 30m of 40mm. And water use always tends to increase over time.

Pressures. I think you'll need two pressure control valves, one on each branch going to the houses. Set them to control downstream pressure so the farthest house has a minimum of 2.0 bars at design flow rate for that branch.

All pipe should be rated for 10bars (8 bars max head plus a little extra for water hammer) to avoid installing pressure relief valves downstream of the controls.

What tank size? Where will you put it?

 

[e331]

Hi, thanks again for you input.
I will check the pipe size and pressure then I get some more free time
Regarding the tank, it's usually sized for 12 hour consumption, however since there is abundant water available, I was thinking that perhaps I don't need a storage tank, just a large well/spring catchment.
Then I found this document which is quite good.
It has sizing requirements for header tank to get rid of possible sand from the water, so I will have to take a look at that.
I will post again when I have found modeled this better

 

[1503-44]

Yes, at the very least, you will need a trash, bugs, frogs, fish, leaf screen and settling tank.
You may also want to include some air valves and flush outs at key locations.

Seasonal usage is also not ideal, as the water standing in the .pipe can go stagnant if it ages, grow bacteria and collect sediment..


Some more suggestions here

https://duckduckgo.com/?q=rural+water+system+design&t=fpas&ia=web

https://www2.gov.bc.ca/assets/gov/e...uidelines_rural_residential_water_mar2012.pdf

https://www.cdc.gov/nceh/publications/books/housing/cha08.htm

 

[bimr]

Just a well known fact. Average usage in the US is 75 to 180gal/pp/d
Google it and you can see 76 to 200 gal/pp/d without opening any url.
I'm OK with yor 100 gal/pp/d number, presuming you have data for your country.

The average water usage in the United States has decreased over the years and is less than those quoted figures. The average water usage is 40 to 60 gals/pp/d. Unless you have some type of lawn watering or other irrigation going on. Also, the average residence typically has 3.5 persons.

Each house should have a service line consisting of a 1-Inch diameter pipe.

The water piping should be sized for peak flow. Peak flow is the hourly flow, not the daily average flow. Peak flow is difficult to determine with such a small system, but is typically 4-7 times the daily average flow.

 

[1503-44]

Average consumption data for anything based on 300MM people probably shouldn't be used as a design value for anything either, but in any case, especially water consumption, there's no substitute for accurate local data.

 

[swazimatt]

You could also consider a break-pressure tank instead of the pressure control valves as it is generally simpler and has less moving parts. This would be a small tank 5m3 (others can give their advise. Have seen them much smaller too) and positioned at N6 would give 4bar at n7

the main pipe would be from N5 to N8 with a tee off to the second tank at N6 (or lower down)this second tank would fill up and when full just have a float valve to shut it off. The pressure will then be measured between this tank and N7 and not N5 and N7.
Could possibly do something similar between N5 and N6 to reduce pressure at N8

 

[itsmoked]

Nice plan. Clean, simple, and provides a lot of safety that can't fail like a PCValve. I like that idea swazimatt!

Keith Cress
kcress -

http://www.flaminsystems.com

 

[1503-44]

Tanks were already discussed.
What's the excitement?

 

[itsmoked]

A smallllllll tank that could even be easily buried.
Though you did indeed mention a tank I was left with the expectation it would be a classic large hard-to-hide tank.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[1503-44]

If no or little storage capacity is needed, it might be be small. With no level control valve to keep from overfilling it, it could be huge. Either way, you need at least one control valve for something. Pressure control, or level.