Centrifugal Naphtha pump install design question

[stewbaby]

Installing a small close coupled centrifugal pump (Goulds NPE 1ST) with single mech seal (John Crane Type 21) in a 5 - 10 GPM Naphtha injection service.

(injecting into suction side of crude charge system , system is at 20 - 30 PSI, to provide additional light ends to cut gravity and help atm tower operation).

Control valve (fail close) downstream of small naphtha pump...if control valve fails closed for an extended period of time before operations catches it...how likely is it that pump would overheat, seal leak, and start a fire?

 

[PUMPDESIGNER]

Would you actually install this pump without some form of no flow protection? I have no knowledge of your application, but I have never in the past, and as far as I can see nevr will in the future install a pump without some form of no flow protection.
We use a simple bi-metallic heat sensor screwed right into the volute. Extremely reliable and cheap.

PUMPDESIGNER

 

[MJCronin]

stew,

Shouldn't your actions and this system evaluation become part of an overall PSM review of the system ? Is OSHA 1910 being implemented in this plant ?

I am aware that many process chemical plants asssign an "engineer du-jour" to modify and rework critical plant systems. Sadly, I am also aware of the "MBA effect" that many plants suffer from...

This effect commonly places "any engineer with any background, to be assigned any problem, to be completed according to the schedule he just developed"

Is that happening here ?

"There comes a time in the affairs of man when he must take the bull by the tail and face the situation." W.C. Fields

 

[stewbaby]

This is in a oil refinery where, industry wide, you'll find alot of centif's without "no flow" protection because system instrumentation (towers & vessels flooding or drying up, flow rate reduction and other) and alarms bring loss of flow to your attention and things run 24/7/365. This one may be a good candidate for a spill back because we want to avoid "over instrumenting" this install.

Yes, this subject is being discussed because of the findings in a MOC (management of change) review meeting.

My question wasn't worded the best...does one typically find, when pump is dead headed, that the temperature increase from energy input by 1 HP motor pump on naphtha at ambient temperature to be worthy of any further thought or ist the temperature rise negligable?

 

[tonyh]

Temperature rise will not be negligible. As a first approximation take no flow power from pump curve, assume no heat loss to the outside, and estimate the amount of liquid in the pump volute. A simple mass times specific heat capacity times Delta Temp = energy input equation will give you an upper bound of temperature rise per unit time. You can then assess the consequences of this.

 

[aberta]

There are certain processes which require canned or sealess pumps for the very reason that a mechanical seal may fail and cause fires. I would not try to justify a mechanical seal. Where temeprature rise is a consideration, we have used a bypass route with a thermostatically controlled process fluid cooler. These systems are with sealess pumps.

Trust it helps you.

 

[smckennz]

Less likely than the mechanical seal failing anyway.
With such a small flow rate, you will probably find that the pump casing will "radiate" the heat stored by deadheading. This depends on temperatures and pressures, but the sums are fairly simple.
The thing that is disturbing to me is relying on a single mechanical seal for fire prevention. The single seal is there to stop leakage, not fires. If you need fire protection as well, you need additional measures.

Cheers

Steve McKenzie