Pumped Network with Pipes Lower than Source Tank 2

[melkordy]

Dear All,

I have searched this forums but couldn't find a thread that describes what am facing exactly so I started this one.

I have a pressurized piping network that is being fed from a pump station and a reservoir that is higher than most of the pipes of the network (the pumps are required in order to deliver extra pressure at withdrawal points and to overcome friction).

When the pumps are not in operation, there is a possibility of having the source tank drain by gravity through the faucets as it is higher than most withdrawal points.

In order to prevent that, I have suggested to force a gooseneck at the pump station just downstream the pumps check valve and the hydropeumatic tank. This is by raising the discharge pipe to a level of 50cm higher than the water level in the tank and place an air valve (vaccuum breaker) at the high point before going down again to the original level and connect to the network.

When power is turned off, flow from the tank (without the pumps in operation) will create a syphon which shall be broken by the vacuum breaker valve and stop the flow.

Is this solution feasible? Is there another alternative?

What precautions shall I consider while designing surge protection for the network (against power failure)?

For power failure, I have placed a typical hydropneumatic tank just downstream the pumps (a bypass line will not work efficicently as the source tank is not much higher than the pump). The problem is that the hydropneumatic tank shall drain completely to the network after power failure. Is using bladder type surge tank more suitable in that case? Also, negative pressures form in the network at high points when the pipes tend to drain by gravity through faucets located at the low points. If I place vacuum breaker valves at thse high points I shall facilitate draining of the network after power failure; Is this acceptable?

Any help will be much appreciated, this is the first time to deal with such situation.

Thanks in Advance
Mohamed Elkordy, M.Sc.

 

[bimr]

Here is another article.

 

[Pumpsonly]

You started off suggesting to install the valve at the pump suction which is against the practice.

Your reply 15 May;

These hydraulicly operated valves made by Cal-Val, OCV and others are intended to

1. Slowly open after pump start;
2. Slowly close at pump shutoff; and
3. Close rapidly as a check valve in the event of power failure
.

extract from the Cla-Val cataloque.
If power failure occurs when valve is open,
the built-in check valve "J" closes immediately to prevent reverse flow.

Extract from Valmatic article page 17
But the rapid closure of either the pump control valve or a fast-closing check valve in a
long piping system poses a dilemma. It was previously calculated that the 8 in. valve must
stroke in 55 seconds to prevent an excessive surge. On the other hand, the valve must close in 5
seconds to protect the pump after a power failure

 

[bimr]

Pumpsonly (Mechanical) posted that "I am not a water hammer expert...".

And so I presented at least 5 independent and separate references above to dispute your water hammer ideas.

Pumpsonly posted “You started off suggesting to install the valve at the pump suction which is against the practice.”

bimr actually posted “Have you considered the installation of an energize to open fail close valve in the suction line to the pump?

How does “in the suction line to the pump” turn into “at the pump suction”

bimr posted “Regarding valve location:

You can install a valve in the suction line or an enhanced check valve (pump control valve), whatever is easier and cost effective. “ and further “His best solution is to install an enhanced check valve”


To summarize,

1. The gooseneck idea is not going to work because

a. The poster will fill up his water distribution system with air which will then be difficult to remove and the air once compressed has a potential to cause water hammer problems in the distribution system.

b. Air release and vacuum breakers are unreliable at low working pressure applications (less than 1 atm pressure) which will be encountered when the pumps and power are off.

c. It is not a good practice to allow air into the water distribution system.

2. His best solution is to install an enhanced check valve where:

a. The check valve will close on power failure to:

1. Prevent siphoning.
2. Help to minimize water hammer potential on pump startup and shutdown.
3. Protect his pump by maintaining water in the pump.

3. A hydropneumatic tank is not practical at a water pumping application where the flows are as high as 330 lit/sec.

 

[melkordy]

Hello All,

Thank you very much for your contribution. I am really proud of being part of this forum as it contains smart people like you..

The attached figure helps describing the system better. I feel that the enhanced check valve will never be able to close after power failure as water continue to flow from the source tank through the pumps.. In order for the automatic check valve to close, the DS pressure should be higher than the US pressure so it can be directed to the valve helmet and push the piston down. Such condition may never happen so I thought of the gooseneck solution.

Air will enter the distribution network anyway; if not from the gooseneck, it will enter from other air valves installed at other high points in the system.. Special precautios should be considered while filling the line again of course to prevent water hammer as you mentioned.. The pumps will always be submerged (no air shall enter the pumps)..

 

[bimr]

You are expecting the air release and vacuum valves ro function at water pressure of 100 cm. Note that air release and vacuum breakers are unreliable at low working pressure applications.

Another option may be to use a float operated valve. This would consist of a float, valve, and mechanical linkage between the valve and float. As the water level drops in the reservoir, the valve will close.

 

[Pumpsonly]

bimr,

1) How high pressure do you expect a vacuum breaker valve to work?

2) When the pump is running , the pressure at the highest point A minimum 20PSIG + 20 meter(29 PSI) height above the tank water level.
Therefore the air valve is seeing 49 PSI. Is this enough to close the valve?

3) New proposed float valve in tank. At what level it is suppose to shut the valve? What would happen if the level drop down to the set level and shut the suction line when the pumps are running?

4)How does "in the suction line to the pump" turn into "at the pump suction"

This only prove how much you know about pumps.

 

[melkordy]

Dear bimr,

I used to know that high quality vacuum breaker valves fully open at a pressure differential of 0.05 bars (negative).. the steady state pressure of the line will be enough to seal the valve of course.

The problem with the float valve you have mentioned is its setting. The water level in the source tank is not constant. If we set the valve to close when the tank is full, it will close even when the pumps are in operation, this will cut the water supply prematurely while the the tank is not empty.. if we set the valve to close when the tank gets empty, it will not close if power failure occurs when the tanks are full and flow will not stop from the tank through the pumps.

Uptill now am convinced with the gooseneck option. The vacuum breaker valve is just needed to break the syphon..

The problem that remains un answered is how to fill back the lines without having excessive transient pressures. The control panel at the station will tend to operate all the pumps in the station as all the pressure is now lost due to power failure. Any ideas on how to do set the operation sequence of the pumps??

 

[bimr]

Pumpsonly (Mechanical): Would have to assume that English is not your first language and things must have been lost in translation. I will not be responding to any of your comments and just take you at your words "I am not a water hammer expert...". and leave it at that.

melkordy (Civil/Environme)

You are expecting the air release and vacuum valves to function at operating pressures as low as 100cm. Note that typical air release and vacuum breakers are not designed to operate and will be unreliable at working pressures as low as this. The vacuum breaker may not open until the water drops somewhat in the source tank. You should increase the 100 cm height to at least a meter.

In addition, point A is 20 meters above the proposed pump. So at the beginning of the power failure, you will initially have backflow conditions from point A to the pump or to the other demand points.

In order to fully understand what you are doing, and develop a reasonable control scheme, it would be helpful to know the following:

Your total flow is 330 liters/sec with 6 pumps. That would make the pump suction and discharge piping approximately 500mm diameter pipe. It that correct?

What is the use of this water system? Is it potable water or irrigation? Note that different equipment will be used for potable water systems than irrigation systems.

Are the demand points neighborhoods or fields? You said they are faucets? This flow is much too large for faucets.

What is the height and volume of the source tank?

What is the control scheme of the water tank? If you install a vacuum breaker, you will not have any control over the height of the source tank. The source tank water level will remain the same once the vacuum breaker opens. If the tank is full, it will stay full. If the tank is low, it will also stay low. The vacuum breaker effectively takes the water source tank out of service.

 

[melkordy]

Thanks bimr,

You are correct, the pipes are 500mm diameter. The network is distributing potable water for municipal and industrial used for a factory area. I meant by the faucets to just describe demand points connected to the network (but they are not really faucets as you mentioned).

As you mentioned, I will have backflow after power failure from point A but it will only last for few moments and the check valve will open again when the nwtwork starts draining. Also I need to install a hydropneumatic tank at the station and connect it to the piping just upstream of the gooseneck. The vessel is intended to decelerate the liquid more slowly after failure. The gooseneck with the vacuum breaker will also function to prevent the vessel from loosing all its water like the source tank.. do you agree with this?

The source tank is 10000 cubic meters in volume (storage for 5 days at peak consumption). The depth of water in the tank is only 3 meters.

As youu mentioned, the vacuum breaker will take the source tank out of service. I would like to add that this will happen when the pumps are off all pressure is lost only. Otherwise, the pumps are operating and keep the network pressurised all the time.

The peak flow of 330 lit/sec requires all 6 pumps to operate. At anaverage flow of 167 lit/sec only 3 pumps are sufficient. When there is no demand, the pumps shall just keep the network pressurized. the pressure felt just downstream the pumps is what will trigger the control panel to operate 3 or more pumps. The problem is that when power is down, pressure shall drop almost to zero at the pumps. This will make the control panel operate all 6 pumps once power is back. This is not acceptable, we should carefully fill in the lines (slowly) to avoid high transient pressures. Any ideas how to set the controls

Your help is highly appreciated.

 

[Pumpsonly]

melkordy,

Pump and motor controls are seldom programmed to auto start upon resume of power supply.
If it is done so, you can add timers to the starter panel of each motor to spread out the starting process over a period that you think is comfortable. If you are using PLC, it is even easier.

 

[bimr]

melkordy

A few remarks on your system.

1. Regarding draining the system.

a. You should be aware that you will experience increased pipe corrosion as a result of draining your water system. The air that enters will cause corrosion and microorganisms will thrive.

b. We never drain water distributions system because it is so difficult to disinfect potable water systems again after the system is drained. You will be introducing contaminants into the potable water system when the system drains.

c. As the linked article notes, it is necessary to fill water distribution very slowly to avoid water hammer, air pockets, etc. The article notes that when filling a water system, air pockets may generate pressures may be up to 10 times normal. This may cause piping blowouts.

2. Have you considered a standpipe? Constructions of a standpipe will greatly simply the control of your water distribution system, especially if you experience frequent power failure or unreliable power supplies.

3. Do you have access to low cost labor? If so, you can get by with manually operated valves. Instead of an automated valve to shut the system down in the event of a power failure, you can have a worker shut the valve(s) manually.

If you system does drain empty, you will have to do a slow startup and refilling of the water system anyway.

You can stop the siphoning with the use of a manual valve closed by a worker.

4. The current practice to control pressure in a water distribution system when you do not have an elevated storage tank is to use variable speed drives on the pumps.

5. Least expensive method to control the forward siphoning..

From the map of your system, one would not expect big issues with water hammer when the pumps fail. The issue with the system draining and refilling will be a much larger headache.

The most inexpensive method to control the siphoning is to install a valve that will automatically (slowly) close on power failure. Such a valve is spring to close, power to open. The valve should energize to open when the pumps are operating.

The valve can be installed and will work in a number of locations, but it is probably best to install it downstream of the pumps.

The valve will probably cost less than the gooseneck and vacuum breaker/air relief on a 500 mm pipe.