Cocker Heater Charge Pump-Seal Plan/Arrangement

[ShahabAPI]

Hello All,
Most of the Cocker charge pumps are supplied by Arrangement-2 of API 682 seal. It means plan 23+52. The experience shows using plan 23 with Gas Oil is super expensive and plan 52 is not a suitable plan in terms of safety and reliability. Another option is to use plan 23+53B, however still we have to use expensive flushing by GO and the issue of cooling during hot stand by condition remain. Does any body have valid running experience with plan 54 without using plan 23? In this case either bac-to-back or face-to-back should be used, however the issue is the process liquid is in ID of seal rings and this is not recommended by API 682, Appendix A. Also, those arrangement have other issues such as risk of O-ring “hang up" and poor pressure reversal capability.
Any experience with plan 53B and "face-to-back" arrangement without using plan 23?

Shahab Zardynezhad

 

[JJPellin]

We have three cokes with a total of six charge pumps. The newest pumps use a plan 54 system with a full skid with pumps, coolers, filters, tank, etc. This plan was worked pretty well. We have had two events where they allowed reverse pressure on the seals which pushed tar into the barrier fluid. We have made procedural changes to try and prevent this. Both pumps share one common system. We are considering adding a smaller system for the spare pump to separate these.

Our two older cokers have single seals that use Plan 32 (outside gas-oil flush) and steam quench (Plan 62). As you noted, this flush plan is outrageously expensive in terms of lost gas-oil. Eventually, we will probably be converting to Plan 54 on these four pumps as well.

All of our seals are high temperature bellows seals so o-ring hang up is not a concern. The seals in the new coker are not balanced for reverse pressure. The inner seal is a stationary bellows with product on the ID. This may have contributed to failures. We are considering design options to address this.

Johnny Pellin

 

[ShahabAPI]

Hello JJPellin, Thanks for sharing your experience. May I know where is you plant location and who is the Vendor for NEW seal cartridges and plan 54? How long are you using plan 54 for the pump? Any failure after modification?
We have three Flowserve cocker pumps (BB2, ~ 13OOkW), originally have been used JC seal with plan 23+52, however we changed one of them to plan 23+53B with Flowserve seal (Face-to-Back arrangement, see Figure 5a of API 682). This change did not solve the cooling issue of the seal during "hot stand-by" and after shutdown. The other option is to add a small circulation pump with plan 53B. However, I personally prefer to use plan 54 without plan 32.

Due to super expensive of using GO for plan 32 (> $300K~1M depends on flow rate, min 8l/min), I am thinking to use plan 54 with "face to face" configuration (See Figure-5c of API 682). Are your new cartridges are "face-to-face"? You wrote The inner seal is a stationary bellows with product on the ID. The issue is: API 682 recommends "face to back" configuration for arrangement-3 because by having liquid on the ID of seal, cause centrifugal forces pushes all particles, dirt, cocks, etc toward the seal rings contact centerline. This is not recommended for this dirty service. I am thinking to use "face-to-back" arrangement of API 682 (See Figure-5a of API 682) and ask seal Vendor to design a deflector to let flush and cool the inner seal and plug the external flush port on the cartridge.
Could you please confirm with using NEW plan 54, you do not use plan 32 anymore.Then, did you experience any issue without external flushing?

We plan to use a common plan 54 skid mounted for all three pumps. Thanks for your time to reply.


Shahab Zardynezhad

 

[JJPellin]

Our Plan 54 system has been in operation for about 5 years. We do not use the gas oil flush (Plan 32) since then. The seals are John Crane. I have included a partial drawing. I would describe this as face to face.

We have experienced a number of failures in the past 5 years. Probably 5 or 6 individual seal failures. They have generally been associated with high heat, plating out the barrier fluid on the faces of the inner seal. Our pumps are two stage, between bearings, approximately 600 HP. We are located in Minnesota, USA. We are redesigning our seals to increase the flow of barrier fluid toward the inner seal.

Johnny Pellin

 

[ShahabAPI]

Thanks Johnny for sharing the information. How many of the six pumps use plan 54 and how many of the pumps use plan 32+52? Do you use a combine plan 54 for multiple pumps? Do you have flushing wearing rings by GO?

Another question is about the suction piping of those pumps. Do you use strainer inside the fractionator? How is the direction of the inlet pipe toward the pump suction compare to the pump shaft? Does the 1st stage is double eyes?
Do you have flashing orifice in the balance line?

Shahab Zardynezhad

 

[JJPellin]

Only two of our pumps are on one common Plan 54 system. The other four are Plan 32 with a steam quench (Plan 62). We have no wear ring flush. We removed the suction strainers that used to be installed right at each pump. There is a Johnson screen in the bottom of the fractionator which serves as the feed drum. Our first stage impeller is single suction, so the direction of the suction line is less important. Our pumps are top suction, top discharge. I think the suction line turns 90 degrees perpendicular to the shaft at the first elbow. The two pumps on Plan 54 have balance lines with no orifice. The other four pumps do not have balance lines at all.

Johnny Pellin

 

[ShahabAPI]

Thanks Johnny, did you calculate how much "GO" will change to cock in the cocker? Assume each seal use 10 lit/min of GO and two pumps running, it means total 40lit.min "GO" is using and during a year (~300Days) the total volume of spent GO will be: 40 lit/min * 60min/1hr * 24hr/1day * 300days = 17,280,000 lit ~ 108,007 bbls. If we assume each barrel of "GO" cost $US30 then the total cost will be $3,240,202.5 per year.
In your case, how much "GO" is changed to cock in the cocker? Do you have any estimation? 10% or 20% or more?

Do you have accumulator in your NEW plan 54? I assume your plan 54 includes: oil reservoir, two screw pumps, two water coolers, one duplex filter, an oil accumulator, piping, valves and instrumentations.

Without wearing ring flushing did you experience any wear or failure? Are your pumps are Flowserve or SULZER? We have Cock fines (<6mm)size and 0.1% wt concentration and sand particles up to 3%wt . I believe flushing wearing rings are not needed as the cock is not abrasive and hard comparing to the hardness of the wearing rings material.

Shahab Zardynezhad

 

[JJPellin]

Our process engineers estimate that 10 to 13% of the injected gas oil flush would crack into coke based on our operating conditions. I used a value of 10% in order to be conservative in my estimates to justify the upgrade. We do not see significant wear-ring wear even without wear-ring flush. As you noted, coke fines are the main solid particles present. They are not particularly abrasive on hard wear ring materials. The flow path through the wear rings passes down along the rotating shroud of the impeller which tends to centrifuge out heavier solids before the fluid reaches the wear ring clearances. You are correct about the components includes in our Plan 54 system.

Johnny Pellin

 

[ShahabAPI]

Hello Johnny-Who is the pump manufacture? Are you using forced oil lubrication or natural lubrication?

Shahab Zardynezhad

 

[JJPellin]

Lots of questions. Pump was Pacific but redesigned by independent shop. Bearings use oil rings. No pressure lube. Ball radial. Ball thrust

Johnny Pellin

 

[ShahabAPI]

Hi Johnny, thanks for your time and sharing your experience.

Shahab Zardynezhad

 

[ShahabAPI]

Hi Johnny,

Do you have pressurized LO system for your Coker pumps? Based on the energy density calculation they do not need, however for pre lubing and post lubing it may be better to consider.

Shahab Zardynezhad

 

[JJPellin]

All of our coker charge pumps use ball radial and ball thrust bearings and are lubricated by pure lube oil mist. Before we added oil mist, they all used oil rings and constant level oilers. No. None of them ever used pressurized lube oil or pre-lube systems.

Johnny Pellin

 

[ShahabAPI]

Hi Johnny,

Thanks for your reply. We use jacketed type bearing housings. Are you using the cooling coil or jacketed type bearing housings? We are experiencing bearing high temperature. I am suspecting to jacketed bearings instead of cooling coil.


Shahab Zardynezhad

 

[JJPellin]

Our housings use pure oil mist, so additional cooling is not needed. However, all of them were originally equipped with cooling coils in the bottom of the housing to cool the oil. The use of a cooling jacket surrounding the bearing is a very bad practice. Studies many years ago showed that removing the cooling water from the jackets can actually reduce the bearing temperature. I would recommend removing the water from the cooling jacket immediately. Put in synthetic oil and run with no cooling or convert to cooling coils cooling the oil in the sump.

Johnny Pellin