Storage of vacuum, possible?

[byl]

Hello,

We are operating a batch reactor process plant.9 reactors. Each reactor is equiped with a feed vessel. We use vacuum to charge them (feed vessel)

At the moment all the feed vessels are on the same vacuum line, connected to a liquid ring vacuum pump.

We assume that we charge about 15-20 feed vessel/day, around 10m3/day.

For practicality reasons, the vacuum pump is left running continiously for only a minimal of use in fact.

I am trying to see if it will be possible to store vacuum in a tank.
This tank will be positionned between the feed vessels line and the vacuum pump and equiped with a pressure controller.
The idea is very simple:
run the pump to reach 50mbar of vacuum on the tank, then turn of the pump,
let the pressure rise to 400mbar as we use vacuum for charging the feed vessel
and then the vacuum pump start again to bring back the vacuum to initial value in the tank (50mbar)

What do you think of this? does anyone having this kind of set up?
One of the issue, obviously, will be lost due to leakage but I think if there is automatic valve positionned on each side of the tanks, that could reduce by a good bit those lost.

 

[Montemayor]

byl:

I've run systems, including batch reactors, much like you're proposing - except that I didn't feed them with vacuum pull and I only used the inherent volume capacity in the vacuum header system. I evacuated and purged with vacuum and my batches were not as many as yours, so I never had a need to evaluate the idea of "storing" a vacuum pull.

The idea is sound when viewed from the aspect of saving the vacuum pump's energy consumption and wear in running 100% of the time, just to have the vacuum pull available when needed. I would also evaluate the application of a vacuum surge drum (or "storage", as you call it) if I were in your position with that many batches to run a day.

The way I would evaluate the application is I would calculate the amount of vacuum needed to bring the reactor (or vessels involved) down to the vacuum level needed to bring in the fresh feed. This is done by calculating the vapor volume in the reactor that will enter the surge drum that is located between the reactor(s) and the vacuum pump. You have to maintain the surge drum down to a lower vacuum then you presently run in your steady system because the vapor entering the surge will nullify a portion of the existing vacuum and activation of the vacuum pump should start at that time. The surge drum size will be determined by how much vapor you're dumping into it and the time you want to save, based on the existing vacuum pump(s) capacity. The surge drum will have to be rated for full vacuum design, so it won't be a "cheap" vessel. The batch time saved + vacuum pump energy & maintenance will have to pay for the installation & instrumentation of the surge drum. If it is economically justifiable, the system should work like a piece of cake.

This is a great exercise in plant engineering improvements and in production economics. I wish you good luck.

Art Montemayor
Spring, TX

 

[Torricelli]

Daer Byl,

This would be theoretically “economical” using a huge vacuum tank with a very small vacuum pump,
like it was sometime done for safety reasons: a vacuum is maintained in a huge tank with a mini-pump (24 hours evacuation time), in case of an uncontrolled pressure increase in the reactor, a rupture disc open to the vacuum tank (volume of tank is 10x volume of reactor).

In your case I would recommend to keep things simple:
- use the system (including the feed tanks) as far as possible a vacuum capacity,
- avoid any leakage by checking how automatic valves are operating
- depending of your vapours install if possible a smaller oil-lubricated or dry running vacuum pump (liquid ring pumps are robust but have a bad efficiency for your aplication and are bad for frequent stop/go).
- If you stay with liquid ring vacuum pump try at least to save water with an air cooled closed loop.

Hope it helps, regards
Torricelli

http://www.vacuum-guide.com