Formalin 37% Pump System

[FACSITEDEV]

Just wanted a reality check on a low-cost low-volume low-pressure formalin pump system we cobbled together using an ITT Flojet 1gpm @ 25psi chemical pump, with bypass shunts feeding back to the pump suction and to the formalin drum.

My thoughts were to start the pump with the drum shunt open until flow was established, then valve down the drum and open the suction shunt so the pressure dropped to the 15psi range required by the formalin drip system at 1-2 gal/hour,
kind of like a hydraulic or diesel injection system, most of the fluid being recycled, and only a small amount used.

I'm using high-precision needle valves on the shunts, and the pump has an internal pressure relief, so I can hold the pressure fairly constant, besides, how precise can Rainbird drip emitters be, anyway?

The existing system uses a pricey peristaltic pump with an overpressure valve rated ±10% and for emergency use only, again, driving 30c plastic Rainbird emitters, kind of like hooking up a 331 Hemi engine to a go-kart drive train.

Any gotcha's I need to correct?

 

[checman]

FACSITEDEV, I am not sure what purpose the return to the suction serves? The return to the drum below liquid level (to keep from entraining air) should work fine.Everything else sounds like it should work great. Returns to the suction can give you some problems depending on how they are piped. 1) Temperature rise, most liquids pick up some heat every pass through a pump. 2) If the return imparts turbulence in suction pipe it can act as an obstruction. 3)Inversely if the velocity head is directed towards the impeller it can increase the pressure on the inlet side which will increase the discharge pressure at a given flow.

Regards Checman

 

[FACSITEDEV]

Because the drum suction runs through a 40-mesh pre-screen, a shunt return to suction is just a way to avoid continous screening of the drum material, albeit only 1 gal/minute.

That is, once the formalin has been screened, might as well return the differential between 1 gal/minute pump supply and 1-2 gal/hour demand to pump suction. Although, as you point out, the liquid will begin to heat up mechanically, I've never heard of any pump smaller than a diesel fire pump casing exploding from recirculated liquid at 25 psi.

It also serves a secondary purpose, if the screen starts to get plugged, having only a drum suction would directly starve the pump and drop the available pressure to the emitters. In an unattended system, the formalin treatment would go off-setting and the operator would have no way to determine what the treatment level actually was, so no way to correct it. Since treatments only last 15 minutes, with the shunt to suction, and a small bladder tank, even if I have only a fraction of a gallon in circulation, my total demand for the full treatment is only 1/2 gallon, so the screen can starve a little and not drop the pressure much.

Anyway, that's the concept. See how it goes in application.
Thanks for the feedback.

 

[Montemayor]

FACSITEDEV:

I don’t know what an ITT “Flojet” chemical pump is (& I doubt if most respondents do) and the description you give is of little use; what we need to know is what TYPE (centrifugal, PD, etc.) of pump you are employing. That information is more useful because it gives us a hint on the discharge and control characteristics of the pump. Is this the peristaltic pump that you mention later on? Nevertheless, since you mention “Chemical pump” and most non-chemical engineers use this generic term to describe a small, positive displacement, chemical addition pump, I am going to assume that is what you are using. If so, then the following applies:

1) Never, but never ever pipe up a pump's recycle discharge line (I assume that is what you mean by “shunt” – an electrical term, not a fluid mechanics term) directly back to the suction of the pump in question. Always pipe the recycle (or “return”) line back to the pump suction SOURCE, not the suction line. Since the pump has to be located near the source due to NPSHa requirements, it follows that it is just as easy to pipe back to the source in order to ensure stable and predictable suction conditions.

2) Use a liquid back pressure control valve (not a manual needle valve) to maintain a constant pump discharge pressure on the pump. This control valve should be installed on the recycle line going back to the source – not the suction. This ensures – for process control and safety reasons – that the pump’s pressure will be steady and constant at the pre-set value you select.

3) From the pump’s constant discharge pressure piping you can now pipe up from this point to the target you have – I presume you are “dripping” the Formalin solution into a chemical process or something similar. For the purpose of controlling the drip (and the pump’s discharge pressure) you must furnish a valve of some sort to meter out the dosage you need to apply. You can either use a manual needle valve (a ½” stainless steel barstock type?) or a process control valve. From a practical viewpoint it will be very difficult to find an appropriately sized process control valve and the needle valve is less expensive and works well if supervised. With the manual needle you run the risk of someone shutting it off – but the back pressure control valve is there to relieve any excess flow rate in the discharge and take it back to the source. I would opt for the manual needle valve as the best practical method to meter out such small flow quantities.

4) Never, but never ever depend on an existing, pump internal relief valve to control your flow requirements. This is, in effect, what you have stated that you intend to do. I take this opportunity to alert you to what experienced engineers in fluid transport already know: positive displacement pumps contain a built-in relief valve for the purposes of an EMERGENCY relief situation – not for flow control purposes. The device is there to protect the pump from mechanical damage due to potential excessive discharge pressures caused by having the pump’s discharge line blocked. I strongly recommend that you install a small, external relief valve on the external discharge piping coming from the pump and send the relief valve’s discharge directly back to the Formalin source – just as I recommend you do with the flow recycle. This is the conventional, safe method to ensure that the process is protected. You cannot rely on the pump’s relief valve for 100% security. This has a standing policy at serious and recognized chemical companies such as DuPont, Huntsman, etc. for many years. Upon doing research, it has been discovered that the pump manufacturers cannot come up with formal calculations and warranties on their built-in relief devices. These items were incorporated many years ago as an empirical solution. They are fine; but they can’t seriously be relied upon from a strict safety point of view.

Formalin solution is very heat sensitive and must be maintained at the pre-determined temperature – otherwise it will polymerize and cause you a lot of troubles. Checman correctly alerts you to this fact and I would add that the way to control the temperature at a constant value is to keep the solution constantly circulating back to the source (not the suction) and only tapping off the net flowrate that you require for your application – much as you’ve described (or I’ve interpreted).

Good Luck.


Art Montemayor
Spring, TX

 

[FACSITEDEV]

Art, thanks for the generic information, it's good advice. To answer your first question, the pump is a diaphram pump. 1. In this case, the needle valve shunt to drum source is in series with the needle valve shunt to pump suction. This allowed a measured %-age of recirculation bypassing the pre-screen, to lengthen the time between screen cleanings in full respiratory gear. With a diaphram pump at 1gpm over 15 minutes run-time, a small temperature increase is more of a concern to us for NPSH-effect, than on any polymerization. But we'll close the shunt to suction to keep you happy. %)2. On your advice, we've located a precision relief valve, made by ValveCheckInc.com that can maintain ±5% at that low pressure/flow. This will be in parallel with the shunt to drum source, so we can tweek emitters drip rate pressure. 3. The client prefers we maintain a precise pressure (and a large manifold), rather than use individual pinch valves on the emitter tubing. See 2 above, ibid. 4.See 2 above, ibid. The main point of this exercise was to show a peristaltic pump is installing a 331 Hemi on a go-kart drivetrain, when a hydraulic- or diesel injection type recirc will suffice, given that the emitters cost 30c and rated ±25% drip rate.