Delayed Coker Unit Heater Charge pumps

[maintennance]

We are in the process of evaluating the offers for the coker heater charge pumps. The MOC of the pumps are complying to C-6 material ( 12% Chrome). Vendors have offered 500-525 KW motors and the pumps are installed with antifriction bearings . The pumps are BB2 type with double volute and one vendor had quoted with diffuser. The mechanical seal flush plan is as per licensor requirement of 32,52.Being mounted with antifriction bearings vendors give accelerometer for off-line monitoring.

Thanks to the users of the coker heater charge pumps to give their valuable experiences on the failure of these heater charge pumps and any modification in the above specification is required.

Thank you onceagain.

 

[JJPellin]

I would have a few concerns. The use of an API plan 32 outside flush is very expensive in this service. In our refinery, we have implemented "deep cut" projects in our crude units to recover as much gas oil as possible. Injecting gas oil flush into the heater charge pumps results in high-value gas oil lost to the coke pile. For one of our units, we estimate that flushing the seals costs us as much as US$1,500,000 per year. We are planning to convert our pumps to a Plan 54 with a system that circulates oil between double seals so that no oil is lost to the process.

Material class C-6 is probably a good choice with one exception. Many of the pump manufacturers will provide class 300 stainless steel vent and drain piping with that class. If you have high chlorides in your feed, you could get chloride corrosion cracking when the right conditions exist during start-up and shut down. We insist on Incoloy 825 piping in this service. I am sure some would consider that overkill. But, a cracked case drain line in this service would result in a major fire.

I see no problem with double volute or diffuser arrangements in BB2 configuration. All of our coker charge pumps are two-stage, radial split, between bearings.


Johnny Pellin

 

[maintennance]

Thank you very much JJ for your valuable inputs. But in our DCU unit pumps, most of the pumps have 32 seal flush plan as per the Licensor document. Any relook required with these pumps with the 54 seal flush plan? Will the mech seal be Face to Face with rotating bellows Seal?

Very good point indicated about the vent MOC.

Thanks to clarify about the seal configuration .

 

[JJPellin]

The coker charge pumps are the only pumps in the refinery that have the particular problem with the gas oil seal flush. In all other services, the gas oil injected is fully recovered in the downstream process. It is only the coker heater charge pumps where a significant portion of the gas oil is lost to the coke pile. I would suggest revisiting the decision to use the Plan 32 flush. Other services with gas oil flush in a plan 32 would be much different.

The charge pump to a gas oil hydrotreater or gas oil desulfurizer would fully recover all injected gas oil. The cost to flush would consist of the energy used to reprocess the same gas oil over and over. The one exception to this would be if the unit was charge limited by pump capacity. In that case, the injected gas oil would be backing out fresh feed which could be quite expensive.

In a coker gas oil pump (gas oil product and/or fractionator reflux) the injected gas oil ends up in the same stream either way. The cost consists of the energy needed to cool it and re-inject it. It should be fully recovered in the fractionator.

I have never heard of anyone using a Plan 32 gas oil flush in the vacuum bottoms service. If they did, this would also cause gas oil to be lost to coke. But, since these pumps are under vacuum, they typically have Plan 54 or some other variation without injected flush.

But, for coker charge service, our experts indicate that 10 to 20% of the gas oil will crack in the heater and be converted to coke and light products. The net loss of even 10% of this gas oil would cost at least US$500,000 per unit per year. I base this on the following: Two double ended pumps flushing 3.5 gpm per seal for 360 days per year at a gas oil value of $30 per barrel with 10% of the gas oil cracked into coke.

If the seal selected ends up being a Plan 32/52 as proposed, I would prefer a standard tandem arrangement with the primary seal having a rotating bellows. I have always preferred the rotating bellows in "dirty" services such as this. We have had some problems with the rotating bellows in this service because of corrosion/erosion of the shell. In two of our cokers, we use rotating bellows and in the other one, we use stationary bellows. With our planned upgrade to Plan 54, we are going to be proposing a face-to-face stationary bellows with the product on the ID of the primary seal and the barrier fluid on the OD of both bellows. Another refinery owned by our same parent company uses this arrangement in their coker charge pumps with great success.


Johnny Pellin

 

[maintennance]

Thanks a Lot JJ for your wonderful and very useful technical reply. Thanks onceagain.