How to select design flow rate for pumps to fill an elevated tank. 6

[termini]

What flow rate should be used when designing pumps to fill an elevated storage tank? Ten States Standard 2007 6.3 references "maximum pumping demand". The average daily flows have been estimated. What peaking factors are recommended for maximum day demand and peak hour demand for primarily residential uses? What is the source of these recommendations? The primary concern is that future demand will be 3.5x higher than the current demand.
Project background: Existing average daily flow = 35,000 gpd; Future average daily flow 120,000 gpd. Flow is primarily for residential dwellings. A 2.5 mile 12" PVC water main extension is proposed to connect to an existing water supply system. A duplex pump station is needed to fill an existing elevated water storage tank. The booster station will be located approximately 1 mile from the elevated tank. Static Head = 100'. Fire water supply is stored within the elevated storage tank.

 

[bimr]

Water demands (average day demand (ADD), MDD, and PHD) may vary considerably between water systems.

The design flow rate for the system supply must be greater than the maximum day demand. The maximum day demand typically occurs during the summer.

Engineers use metered use records to quantify the ADD for most water systems. Daily source meter records provide an accurate estimate of MDD for the water system. Note that daily meter readings may result in inaccurate MDD estimates if the water system operates sources intermittently, does not collect meter readings at the same time every day, or does not include changes in storage volume over time.

The MDD (Maximum Day Demand) to MMAD (maximum month average day demand) peaking factor typically varies from 1.3 to 1.7.

 

[Lan123]

What state is this? You may also check the state EPA water storage regulations.

 

[cvg]

back up just a step
it all depends on the way you tend to operate the pumps and tanks. It may or may not be related to the MDD. Depends on the number of tanks, volume of tanks, number of pumps / wells etc. The information you have given is not adequate. Your last sentence indicates the elevated tank is for fire flow. Is that all it is for?

If you plan to replenish tanks at night with pumps shut off during the day, then your pumping rate should be sufficient to completely fill the tank(s) during approximately 8 hours at night with only night time demands drawing from the tank(s). If your pumps are also supplying part of your daytime demand, then you need to look at full time operation.

Your peaking factors are more useful for calculating pressures during water delivery, not as useful for calculating a pumping rate. Your combined pumping and storage / transmission system should be able to deliver ADD, MDD, PDD, PHD and FF all at the required minimum and maximum pressures as well as handle fire and replenishment flows.

 

[stanier]

You model this in Epanet and ensurre that your tanks are sized adequately and the duties of pumps established.

"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/

 

[termini]

Thank you for your input. The project is located in Maryland. The pumps will be reviewed by Maryland Department of the Environment. The pumps will turn on based on water elevation within the single 200,000 gallon elevated storage tank. Therefore the pumps will be in full time operation. The tank is to hold domestic and fire flow. (fire flow pumps are not planned) The water supply is from another private water company's distribution system that is able to provide the expected demand. I would like to determine a design pump rate to select adequate pumps that would handle existing demand and consider the future demand. What drawbacks exists if the pumps were designed for the future flow demand (3-10 years)? Would Variable Frequency Drive Pumps be beneficial to handle future growth? What basic calculations may be run to estimate the pump size prior to creating the water system computer model?

 

[bimr]

You should reconsider the fire flow scenario since you probably do not have an adequate volume for the fire water supply. A typical modern fire water flow would consist of 1,500 gpm for 2 hours. That is almost 200,000 gallons. The flow water supply requirement is typically established by the state insurance board.

You may have to have a larger water supply since you probably do not have enough fire water storage.

You need to obtain the existing water pumping flows records. One would expect that the water consumption in the summer will be the critical time. You probably want to size your pumps to supply 1.5 times the maximum month average day demand. The maximum month is probably July.

In the beginning years of the pump system operation, your pumps will probably operate for fewer hours per day. Over time with increased capacity requirements, the daily operational hours will increase. There is no drawback to operating the pumps periodically.

Variable Frequency Drive Pumps will probably not be appropriate since the pumping head will not change.

 

[waterpipe]

1. Regarding your question about the source of recommendations, here are some books (handbooks) that contain the full coverage of your problem:

- Water Distribution Systems Handbook- Mays
- Twort's Water Supply
- Advance Water Distribution Modeling and Management- Haestad methods

2. In the line of cvg reply, it depends on your functional design. It is in relation with your tank sizing. I'm not in state but it seems that each state has its own outlines for tank sizing. Some options include the following:
- operate at a constant rate
- adjust flows to roughly match the demand (minimize use of storage)
- pump during off-peak electricity pricing
- match to demand by using variable speed pump
- looking to permissible number of on/off

3. However, you said that pumps will be in full time operation. My suggestion is to use constant speed pumps and size so that you can add another one in the future when demand increases. Do not design them for the future consumptions as there are normally uncertainties over future demand estimates.

4. And running Epanet is easier than it looks. I recommend it even for a pre, pre, per-feasibility calculations.

5. BTW, the future demand of 3.5 times higher seems odd to me (if you've referred to liter per capita). I assume you have meant the total volumetric consumption that includes the population growth.

6. And what does it means "A duplex pump station"?

 

[cvg]

duplex - two pumps
1 - 100% and 1 - standby?
or triplex
2 - 50% and 1 - standby would be better and would handle future growth better. And put in a fourth spot for a future pump

agree that you should already have a working model at this point, it's just not that difficult to do

 

[77JQX]

I agree with bimr about the fire storage. You say the uses are primarily residential - which implies there will be other fire flows to consider. In my jurisdiction, the minimum commercial or school fire flow is 2,000 gpm for two hours, and goes higher. I'd find out for sure the maximum fire flow. You should also consider how you'll provide fire flow while the tank is out of service for recoating.

As for a VFD, I wouldn't consider it. With a tank, you'll be able to select a single performance point for your pumps, and cycle the pumps as tank levels vary. With a pump operating at BEP, that's the easiest from a control standpoint, and also the most efficient. With a tank, there's just no reason to mess with a VFD.

I think the important thing to consider is your transmission line. I like to shoot for a friction loss of around 5' per 1000 feet of pipe as a reasonable compromise between pipe cost and pumping cost. That would be about 1,500 gpm in your 12" PVC pipe. You'd have to determine whether a flow that high would result unacceptable pressures in your system, both on the suction and discharge side of the proposed pump. 1,500 gpm is also overkill for your current tank size, but I don't think I'd go below 1,000 gpm. Sure, smaller pumps are cheaper, but the actual pump cost is a small part of your pump station cost, and smaller pumps wouldn't have significantly less friction. Two 1,000 gpm pumps may also meet future fire flow when the tank is out of service. (In my jurisdiction, a genset is required at a pump station when the pump station is used to provide fire flows.) If the pump station could be used to provide fire flows, you might avoid having to build a larger storage tank. Using the pump station for fire flows may also allow the utility to draw the tank down and take advantage of off-peak electric rates to fill the tank (as cvg has suggested).

Peak factors for maximum day and peak hour are highly localized. I doubt you'd want to use mine, because we're in the desert and our residential use can include a substantial irrigation component. Check with the local purveyors.

And that's my $0.02.

 

[stanier]

Once you have selected the pumps and designed the system you can consider the waterhammer analysis. Things to consider with PVC pipe are:-
a) The fatigue levels for stopping and starting pumps derate the PVC design life

http://www.pipa.com.au/index.php?option=com_content&view=article&id=54&Itemid=72#pop101

b) If fire hydrants are in use you may find a quick closure valve on high flow will give a high transient pressure.
c) With a low surface coefficient you will find the transient pressures take longer to decay.
d) The low celerity of a PVC systems means the transit time 2L/a is longer and thus valve closing times needs to be longer
e) Joukowsky does not predict the high pressures resulting from column separation/rejoining . These may be greater than predicted by Joukowsky
f) A VFD may be useful in reducing fatigue in stopping and starting but will have higher embodied energy and a larger carbon footprint. See postings on VFDs in Eng Tips



"Sharing knowledge is the way to immortality"
His Holiness the Dalai Lama.

http://waterhammer.hopout.com.au/