Downhole Chemical Injection panel

[anji71]

long one i'm afraid...

I'm trying to inject corrosion inhibitor downhole to each subsea well. The downhole pressure of where i'm injecting to is 60bar u/s of a ESP pump. I have a pump topsides where the chemical is distributed to a subsea flow distribution panel to inject to individual wells via controlled Amflow valves( valve provides constant,pressure independant flow of injection flow).

In theory as long as he pressure on the pump topsides(pressure controlled via a back pressure regulator) exceeds the downhole pressure( set at 80 bar) then i should be able to inject the required flows to each well. However on some of the wells I get a back pressure of 40 bar on the indicators on distribution panel, and some with 0 pressure although all are at 60 bar downhole pressures.

The problem is some of the injection points at the panel lose some flow completely,then re-appear with no known reason.
I find that if i set one of the rates on the amflow valve too high( 5L/min as opposed to 2L/min) then this can reduced the flow on one of the other injection points even though they are completely independant.

Even though the pump is set to deliver over 500mls/min the actual daily injection rate is only 250ml/min, with the flow meters on the distibution indicating at the correct treat rate volumes.

Any Ideas please?

 

[injectionguru]

Dear anji71:

An even longer reply; I'm afraid...

The problem that you described is commonplace; take the following into consideration:
1) I am not familiar with the control device, although I assume that it is an IRCD (injection rate control device) consisting of a biasing regulator (biased to downstream pressure) with a metering valve that is either integral to the IRCD itself or installed in the tubing or pipe downstream of the IRCD outlet. No matter, the theory of operation is that that the IRCD maintains a constant delta P across the metering valve and, as such, the injection rate is a function of the selected (adjusted) Cv of the metering valve.
2) Critical conditions for proper operation and injection rate stability are: (a) the inlet pressure to the IRCD must be very stable. Pressure stability can / will be greatly influenced by pump stroke volume and / or speed (strokes per minute); the extent to which the discharge pressure is dampened (+/- 5% of discharge pressure peak-to-peak is recommended); the deadband of the backpressure regulator and, the deadband of the IRCD. It is assumed that the backpressure regulator and the IRCD are spring-loaded devices. (b) the fluid must be very clean; the smallest of particles in the fluid stream can / will exert a profound influence on the actual versus adjusted (desired) Cv of the metering valve; the lower the injection rate, the more critical fluid cleanliness becomes. Cleanliness on the order of NAS 6 to NAS 8 is recommended. (c) line and component sizing, i.e., the inline components, tubing, fittings, check valves, ect. must be properly sized taking into account, in particular, the specific gravity and viscosity of the fluid. Viscious fluids are very hard to control using IRCD's...
3) All that said, check / do the following: (a) assuming that you have more than one injection point, take the highest backpressure of all injection points and add to that value the value of the IRCD spring, plus estimated system pressure losses and set your backpressure regulator to relieve at the sum of those values. For example, let's say that you have an injection point where the topsides backpressure is 60 barg and the IRCD has a 14 barg spring and it takes 3 bar to overcome the spring in your check valve (assuming that you have one downstream of the IRCD / metering valve)... and line losses in the tubing from the outlet of the IRCD to the departing bulkhead fitting on the panel are estimated at 0.5 barg... it follows that your backpressure regulator should be set (to begin) at 60 barg + 14 barg + 3 barg + 0.5 barg = 77.5 barg minimum initial backpressure regulator setting. (b) take the sum of all single-point injection rates and multiply that value by 1.15 - 1.2 and set your pump to deliver not less than the resulting value. For example, let's say that you have three injection points, each of which need 0.50 l/min... it follows that your pump should be set to deliver 0.50 l/min + 0.50 l/min + 0.50 l/min = 1.5 l/min x 1.15 - 1.2 = an initial pump rate setting of 1.725 - 1.8 l/min.
4) The above done, this is the point at which discharge dampening and deadband characteristics in the backpressure regulator and the IRCD itself become all too important. But, assuming that the inlet pressure to all IRCD's is now very stable, i.e., holding at 73.6 - 81.37 barg... and... assuming that your pump is delivering 1.7 - 1.8 l/min... and... assuming that you have a very accurate flowmeter... you should be able to adjust (slowly) your metering valve (s) to the desired injection rate (s) without one affecting the other and, the injection pressure to each point (s) also should be very stable.
5) On the injection pressure front... bear in mind that the pressure that you read topsides will be lower than the pressure downhole due to static head. That is, if you have a downhole pressure of 60 barg, that does not equate to needing to overcome 60 barg topsides. The topsides pressure at the injection point connection on the tree or umbilical is the value that you should use to determine the intial backpressure regulator setting as described in 3) above.
6) If you observe pressure on the inlet side to the IRCD (s) that is lower than the initial backpressure setting or, if the pressure swings between the desired backpressure setting and a lower value, it is likely that your pump is undersized or, the range of the deadband in the backpressure regulator is too wide for the application. Diametrically, if you observe pressure on the inlet side to the IRCD (s) that is greater than the initial backpressure setting or, if the pressure swings between the desired backpressure setting and a higher value, it is likely that there is insufficient dampening on the discharge leg of the pump or, again, the range of the deadband in the backpressure regulator is too wide for the application.
7) If the pressure on the inlet side to the IRCD (s) is stable, but the pressure downstream of the IRCD (s) and the metering valve (s) swings high then low, it is likely that the spring in the IRCD is not the correct range for the application or the piston in the IRCD (s) is sticking due to friction... then suddenly breaking loose once the downstream biasing pressure increases enough to overcome the friction. That is, if your injection pressure is "nervous" or it "chatters" ... this behavior suggests a variety of problems. Things to look into if this is the case include: (a) the IRCD spring range, (b) the cleanliness of the fluid, (c) the elastomer (s) in the IRCD, particularly the IRCD piston O-ring (s)... if the elastomer material is not compatible with the fluid, it can / will become too hard or too soft to do its job or, (d) the check valve poppet or ball could be sticking then breaking loose to produce the same behavior.
8) All of the above taken into account and looked into... if you still have a problem, e-mail your P&ID's to me and I'll try to help. My e-mail address is: monte_goodner@w-industries.com I've been designing and building injection systems largely for deepwater projects for about eight years, so I've come across virtually every problem one can imagine. No fees... no selling... I do a lot of study and trouble-shooting work for producers and engineering concerns.

Regards,

injectionguru