Tank farm commisioning

Hi guys. Procedural commissioning question here. I know in the proces industry it's pretty typical to do a wet test, or water run during commissioning, but I'm not sure if the same applies to a tank farm system. The tank farm is storing and distributing primarily oils, which is why I'm not sure that a water test is a good idea. Looking for thoughts and opinions.

Thanks.

Russ

Hi the question is a bit vague tbh. You need to clarify things like type of sensor etc. For example a capacitance rod will need to be checked with the correct die electric fluid where as a differential pressure sensor or load cells dont care....

provide more details and you will receive a better answer

Define test purposes please.

Dan Bentler

I'm writing specifications documents and defining the commissioning phase of the project. I'm thinking it would be nice to actually pump water to test piping, pumps, etc. However, since the system will be distributing oil, I'm not sure introducing water right away is a good idea, and leaning towards simply doing a dry test of appropriate equipment and then actually finalizing the comissioning with actual oil. Just looking for thoughts, in case someone has had a terrible experience skipping a water run and going straight to the actual fluid.

Russ I think it comes down to this "what is cheaper and easier, water leak or oil leak?

If you test with water -- WHEN you get a leak -- you do not have to file a bunch of reports and you can let it go down drain, plus risk of fire and a mess to clean up.

If you test with water then you will have to separate water from oil in early startup. I think the cost of this is a lot less than the potential cost of a large oil leak.

You could test with nitrogen or even compressed air but the risk here is component failure and shrapnel. DO NOT do this with plastic pipe. Also you will have leaks that you will not have with a liquid - pump shaft packing for example. The other downside is you are not able to test with pumps running and find pressure surge problems or cavitation problems.

Think I would go with water and plan on doing oil and water separation.
Dan Bentler

It is hard to give you an exact answer to this question. You should really discuss this with the different people who are involved in the project (including the customer!). How much will it cost you when you need to remove the water after the commissioning tests? And the more important questions, how much will it cost you if you start with the real product and something goes wrong (product loss, dangerous situations, impact on the environment,…).

Normally the right answer here would be to ask the client and let them dictate. The problem is that they are largely unresponsive. The project on their end is handled by senior level people, and the impression I get is that they are leaning on us to just handle the details and give them what they need.

In light of leitmotif's response, I am leaning towards including a water test, and planning to separate water from the oil as a step in the final commissioning. At least if this is called out, it will be a cost that is identified during bid and not a change order.

Thanks guys.

What type of products are these tanks containing?
Are they pressure vessels, floating roof cylindrical, etc?
Who will be constructing the tanks?
Will there be water pressure testing of the associated piping?

Generally, the people that build the tank will test or have it tested for integrity. Maybe the owner. That may be done at a earlier time. That test can be a good time to check your transmitters. Unfortunately, cable runs probably won't be done by then. That might be the only time you will see water in the system. They might use the same water and go tank to tank as they are built. saves on treatment costs.

Some of the tank farms where I worked had millions of barrels of storage, so a full water test would probably put your bid too high. The commissioning will likely be:

1, Check process loops(cables)
2, Calibrate transmitters
3, Bump motors for rotation
4, Check discrete valve operation and feedback
5, Check control valve operation and feedback
6, Tuning PIDs
Etc

Software/SCADA/HMI should be really close(95% plus) at FAT. Valve interlocks, PID direction, alarms, shutdowns, permissives, etc.

Enjoy

SD_Scott,

I have all of your indicated actions accounted for. However, my concern is not with tank integrity, but more with piping and distribution itegrity. Should all of the activities you outlined be successful and oil is introduced to the system for distribution, we could run the risk of a large oil spill. In light of that, cost or not, I think it's wise to plan for a full water test. I was checking here to see if there was huge consensus against a water test, but there appears to be none, aside from your post indicating it is costly. At least if it is a known cost, it can be accounted for.

Thanks to all for the input.

russrmartin said:

Hi guys. Procedural commissioning question here. I know in the proces industry it's pretty typical to do a wet test, or water run during commissioning, but I'm not sure if the same applies to a tank farm system. The tank farm is storing and distributing primarily oils, which is why I'm not sure that a water test is a good idea. Looking for thoughts and opinions.

Thanks.

Russ

Click to expand...

I understand your concern. We are in the process of commissioning a large ice cream facility from the ground up. Chocolote enrobing products have the same issue, no water can mix with the chocolate, so we have accounted for during our testing ample time for all tanks and pipes to dry out before any chocolate can be received.

You have to test on water, as commissioning with product is always an expensive alternative.

Pressure Test??

Russ,

Are you able to to a pressure test?? I have not worked on a oil tank farm system before, but would that not test the integrety of the unions and welds?? Maybe with nitrogen?

Just a though...I'm probably out to lunch....

Andew

Andrew Evenson said:

Russ,

Are you able to to a pressure test?? I have not worked on a oil tank farm system before, but would that not test the integrety of the unions and welds?? Maybe with nitrogen?

Just a though...I'm probably out to lunch....Andew

Click to expand...

At Trojan the containment was pressure tested to 60 psi using compressed air. Air was provided by 4 or 5 LARGE diesel compressors pumping to an 8 or 10 (?) inch manifold. Took 48 hours to pump up with compressors running near constant.

I think air much less nitrogen is cost prohibitive for million gallon tanks.

Dan Bentler