Best Practices for API Plan 32

[Rotologo]

Hi community,

I just joined this forum and I was hoping to get some information / advise on what are you guys doing out there when it comes to operate / preserve mechanical seals having API Plan 32 as primary support system. The problem I'm facing as follows:

Operators are restricting the flow towards the mechanical seals on those pumps in hot standby, I have found out they do this for two reasons: to minimize the flushing consumption and to prevent the pump to cool down and seize (on the throat bushing area) as apparently this has happened in the past in some pumps. This have caused several mechanical seal failures during start up when the standby is needed for operation. After opening the seals I have seen a lot of contamination and in some cases even corrosion on the seal head which indicates a less than adequate flushing supply. Questions;

1.- Do you keep the same flow to the mechanical in a running pump and in a hot standby pump?. I know that seal chamber pressure will be different when the pump is down (especially in OH pumps) but what variable you control when the pump is not running, do you try to keep the recommended 15 ft/sec or the 10- 20 PSI above the seal chamber pressure in the hot standby pumps as well?.
2.- Are you permanently monitoring the seal chamber pressure (whenever there are provisions in the seal chamber to do that?). I have considered to start this practice but there are some concerns on creating new possible locations for loss of containment (hazardous fluids).

These mechanical seals are installed in a Delayed Coker and Vacuum units so in most of the cases we are talking about HC @ 500 - 700 *F and with relatively high particle content. The flow control is achieved by restriction orifices (supposedly) calculated to provide 15 ft/sec to the seals.

I'm new to this installation and I know there is a lot of education to be done in terms of understand how to operate the seal supporting system as well as in correcting design deficiencies (flow meters, warm up lines, etc.) but for now as I stated before I'd like to know your experience in the operation of the seal supporting system on hot standby pumps.

I'd appreciate your comments.

Regards

 

[JJPellin]

The use of Plan 32 can be very expensive for pumps in these units. In our crude units and cokers, we commonly used Plan 32 with HCGO or HVGO as flush. We have determined that flushing the coker charge pumps in particular can cost millions of dollars (US) per year. We are in the process of converting these pumps to Plan 54 systems.

For the pumps where we still use Plan 32, I can provide some answers to your particular questions. First, we do normally keep the gland oil flowing to the off-line pump. Most of these pumps are set up to start automatically. We adjust the flush using flow meters rather than orifices. The flush flow rates tend to stay fairly constant with the pumps not running for most of the pumps. The exception would be two sets of our coker charge pumps which are two stage pumps that do not have balance lines. So, the seal chamber pressure on one seal is drastically different when it is not running.

Our operators regularly turn down the flush flow rate on the off-line pump. They do this for a few reasons. First, we do not have an overabundance of flush available in some areas and need to minimize the flow to off-line pumps in order to have enough for running pumps. Second, the flush is generally lighter than the pumped product and can tend to flash as it enters the product stream. This can upset the running pump. In addition, if the pump case fills completely with seal flush, it can cause problems when the stand-by pump is started up. The slug of lighter product can flash in the line and register as a loss of flow on the heater-pass flow meters. This triggers an automatic heater trip, velocity steam or emergency steam in some systems. If the flush flow is kept low and the warm-up flow is sufficient to keep it diluted, this problem can be avoided.

We do not use continuous monitoring of seal chamber pressures on any of our pumps.

We set our flush rates based on flow, not pressure or velocity. The flow rate target is determined by the seal manufacturer for the given seal size, speed and design.


Johnny Pellin