Suction Head of PD Pumps

[Ali007]

Hi Folks

Wondering if you can help. We have a small chemical injection PD pump. We are finding that it is impossible to control the pump and we are injecting far too much chemical.

Is it possible that the chemical is going straight through the pump. Is it possible that the suction pressure is too high and is overcoming the pump? Is it possible the velocity in the line is too high?

I am aware that the NPSHa must be greater than the NPSHr but is it possible to have too much NPSH avialble?

Any comments or advice welcomed.

Regards

Ali

 

[JJPellin]

We have many installations where we use small PD chemical injection pumps to meter the introduction of chemical. In some of those installation, the suction pressure is greater (because of the liquid level in a supply tank) than the discharge pressure. Some of these are discharging into an open sump, for example. The pump (if it is a diaphragm or piston pump) uses two check valves to control the passage of fluid into and out of the pump. Even if these check valves are spring loaded, they cannot be considered as tight shut-off. If the check valves have any damage or debris, they can leak by, allowing the chemical to free-flow through the pump. This results in excess chemical addition. There are a couple of solutions that we have used with great success.

First, the pump manufacturer probably has a recommendation for a minimum back pressure. If the pump does not have enough back pressure to re-seat the check valves, it will not provide accurate delivery. They usually sell back-pressure valves just for this purpose. They are basically pressure regulators that hold back 30 or 40 psi so that the pump has stable back-pressure against it. This improves the accuracy of the pump and avoids the problem that I described above. This is the preferred solution.

In some cooling tower applications, we simply extended the discharge line to an elevation higher than the top of the supply tank so that there can never be higher suction head than discharge head. The discharge is then introduced to an open line or trough that allows the chemical to free-flow down to the cooling tower sump. You should note that it does not work to just run the discharge line up to elevation and then back down. If you do this, it will just siphon and you can still have the same problem. You have to have a vent at the top to break the siphon.

But, of these choices, the back pressure valve is by far superior. It avoids the over addition of chemical and greatly improved the accuracy.


Johnny Pellin

 

[Compositepro]

Most chemical metering pumps have spring loaded check valves. They are designed to seal pretty tightly - otherwise they could not meter accurately. If your suction pressure is higher than the discharge pressure it is very possible to have uncontrolled flow through the pump. Some pumps are not designed for positive suction pressure. (Most double diaphragm air pumps, for instance. It causes reversal of diaphragm flexture on every stroke which will quickly lead to fatigue failure).

 

[MartinSr00]

On diaphragm pumps, the two check valves pass flow in the forward direction and prevent it in the reverse direction.

If your suction pressure is higher than your discharge pressure, by an amount that exceeds any spring pressure in the valves, the flow will flow straight through the two valves no matter what the pump does. Essentially, you've plumbed to check valves in series and pushed flow in the unchecked direction.

I don't know what chemical you're feeding, but if the chemical isn't too exotic and is relatively clean, magnetic drive gear pumps work fantastically. We use them for water treatment chemicals where I work. They're available in exotic metals (titanium, alloy 20, stainless, etc.).

If your turndown rates are not high, simply vary the pump speed. If they're very high, put a valve in series with the discharge, and a relief valve piped back to suction set at a reasonable pressure. Then just open or close the valve, and the flow meters almost perfectly. Smooth pulseless flow.

 

[pleckner]

Add a back pressure regulator, the self contained (cheap) type should be sufficient. You can get the minimum set pressure from the vendor.

 

[dabluffrat]

what type of PD pump? In piston or plunger type pumps the NPSHr is basically the pressure required to overcome losses across the suction valve and lift the valve spring. you can increase the spring stiffness to increase NPSHr slightly, but it sounds like your spring may be too weak, or already broken, and the high suction pressure keeps the suction valve always open.

Did you know that 76.4% of all statistics are made up...