Here are few things that ate our lunch on an old Girbitol Mea Unit on Low Pressure H2 plant.
Small air leakage on pump seals can be a big problem.
We used a mechanical filter (Cuno)prior to the carbon towers.
You have to have carbon towers (twin) with the proper grade carbon (liquid/liquid). Using 3-way valves, one was always on line.
We used an antifoam, not a silicone. If I remember correctly it was an "Igepal".
All our exchangers were 316 SS shell and tubes.
The main thing that helped alleviate the corrosion problem and pluggage of the HE's was a midnight redesign project of the tube layout on one HE. The existing tube bundles were on a triangular pitch with all the possible tubes you could cram into a tubesheet. We built our own tube bundles and as we had gone through all the permutations of tube materials and sizes, etc. The problem was that with a triangular pitch which made it impossible to clean the bundles or them to stay clean for any length of time. Myself and the fab shop technician decided to build a square pitch bundle for one of the worst offenders. It was built and installed and after 6 months was pulled and no corrosion nor fouling, the tubes were shinny. We got caught during startup by a sharp process engineer who had been assigned to find out why one E-7 bundle on one train was working like a champ and the E-7 in an identical system wasn't doing well. What we lost in theoretical heat transfer, we more than gained in operational performance.
All of the troublesome bundles were changed to square pitch within 6 months and all ran for 2.5 years until the plants were replaced by a single H2 plant with a PSA purification system.
