Consideration for Water Pump repeated switching

[I_am_learning]

I am designing a digital circuit for water pump control to always top-up the over-head tank by pumping water from the sump. I am using sensors in the over-head tank.
I am now wondering why shouldn't I start the motor as soon as the water level in the over-head tank drops just a little below the top. Yes this results in the pump switching rapidly, but the total running time remains practically the same.
What problems may switching the pump too often create apart from the wearing of the relay used to switch it. Since the motor used in the pump is simple induction motor, I don't see how it can be harmful for the motor to be switched too often.

The advantage of this scheme is that the water levels being at top provides higher pressure, and more importantly, more backup water in case the system fails or there occurs power outage.

 

[Averagesupernova]

There is more stress on a motor when it is started than any other time. Just because it is an induction motor doesn't mean it is immune to this. I would not rig it so that it starts just to top the tank off. I would allow longer run times with less starts.

 

[I_am_learning]

OK I count that, It increases the likelihood of the stater coil falling out. But How much likely is that compared to the life of the motor. I mean should that really come into consideration?

As for the relay, I am searching for something more reliable, may be a Solid-State-Relay?

I am really looking forward for other factors to be considered?

 

[sophiecentaur]

If you want to limit the amount of pressure variation then you could have a bigger tank - so there would be fewer switching operations. Or you could pay more money and buy a better specified motor.
I don't know what the actual requirement is but you could use two pumps, one to provide smoothing and another to kick in when there is a heavy demand.
Or, two other approaches:
1.Have a flow rate control valve to control the supply rather than rely on the constant head from your tank.
2. Have a second constantly running pump which tops up a header tank which spills back into a reservoir just below it (like diesel fuel systems in motor cars). There's nothing like a weir to define a head of water over a large range of flow rates.

 

[I_am_learning]

Hi sophiecentaur, thanks for your reply. My main aim was to always top-off the water because, power-outage is frequent and loooong here.
And I was asking about whether I can switch on/of regular water pump frequently, i.e. re-top the tank as soon as the water level reaches say just 3" below the top? It won't result in extra electricity bill I think.

The pressure thing just came to my mind and isn't that important.

 

[Averagesupernova]

Once again, the final scheme depends on the requirements. Naturally, how else would it be? I didn't know you had frequent and long outages. This changes things. You just may want to consider the possibility of a shortened life on the motor a cost of reliable water supply. Look into a motor that will take frequent starts. Not sure what type of pump you have but maybe there isn't a lot of startup load.

 

[sophiecentaur]

Or decide just what range of pressures you can put up with. If pressure is not such an issue then you can obviously afford a bigger range of levels and, consequently, switch the pump less frequently. It all depends upon the specific circumstances. If you are just finding the pump is witching on and off too fast then build in some hysteresis in your level sensing feedback.

 

[I_am_learning]

Or decide just what range of pressures you can put up with. If pressure is not such an issue then you can obviously afford a bigger range of levels and, consequently, switch the pump less frequently. It all depends upon the specific circumstances. If you are just finding the pump is witching on and off too fast then build in some hysteresis in your level sensing feedback.

why shouldn't I switch the pump frequently? I am tired of repeating the same question.

 

[Mike_In_Plano]

It seems as though you have a classic engineering problem - compromise. As long as values and / or actual experience are left out of the decision making, any idea can be rationalized.
When very little supportive evidence is to be had, looking at the extremes of the problem can prove helpful.
For example, given a 20 gallon tank, would a cup of water represent an excessive burden on the user should the power be removed. Well, 1 cup represents .3% of the system's reserves, so I think that most people would judge refilling the tank after every drink as excessive. I suspect the same would hold true for any small percentage, i.e. 2-3%.
The more difficult question arises when you're talking about the impact on quality of life. When 10% of the tank is drained, or even 20%, you can expect people to notice.
The art of engineering is putting aside the rationalizing aspect of our mind and substituting those values we can measure or deduce from observable behavior.

Best Wishes on Your Project,

Mike

 

[sophiecentaur]

why shouldn't I switch the pump frequently? I am tired of repeating the same question.

Only the manufacturers could tell you about that with certainty. Have you told us the actual rate it's going at the moment? Once a second? ( clearly not good) Once in ten minutes( sounds ok to me). I don't have a more definite answer but I can suggest some ways of increasing the time interval.

 

[I_am_learning]

That will be like once in 8-10 minutes during the peak demand (of water).Thanks for all of your responses. I will try it. After all even at the worst there is just a water-pump (and few relays) to loose in exchange for a great practical experience !

 

[sophiecentaur]

From that reply, I shouldn't think there would be any problem at all. With all that time between startups, I shouldn't think you'd have any overheating. But contact the manufacturers for info.