PDA

View Full Version : Pressure testing



Chopper
23-10-2005, 07:31 PM
Has anyone ever heard of pressure testing a large ammonia plant by blowing the pressure up to approx 8 bar with an air compressor and "topping" off the final push with ofn? Is there anyone out there doing the same? Am i the only one that see's this as pretty stupid?:confused:

US Iceman
23-10-2005, 11:35 PM
I would rather see dry nitrogen used for the entire pressure test. However, even using compressed air, the system should be evacuated to remove all non-condensables (air or nitrogen).

I have seen both of these done before, for leak checking. Even after the leak testing, and after ammonia was charged into the system, small leaks can still be found.

Typically, some small leaks will develop through thermal cycling of pipe, fittings, or flanges. Or, leaks will also show up where the air or nitrogen would not.

Any moisture that is admitted into the system using compressed air will be absorbed by the ammonia, so I don't think I would get excited about a little water.

I would be more concerned about any non-condensables being left in the system.

The compressors should not be used to generate the pressure for testing. This might seem like a common sense statement, but sometimes common sense is not so common.

Argus
24-10-2005, 12:50 AM
I would rather see dry nitrogen used for the entire pressure test. However, even using compressed air, the system should be evacuated to remove all non-condensables (air or nitrogen).

Any moisture that is admitted into the system using compressed air will be absorbed by the ammonia, so I don't think I would get excited about a little water.

I would be more concerned about any non-condensables being left in the system.

The compressors should not be used to generate the pressure for testing. This might seem like a common sense statement, but sometimes common sense is not so common.


I agree with that except that I?m not certain that I would want to introduce oxygen into a system containing air.
The critical mass in air is probably too small to cause oxidisation but I wouldn?t want to take that risk.

OFN is safest. There have been some accidents in the UK lately with systems under pressure and the HSE (Health & Safety Executive) is very concerned about it.






Typically, some small leaks will develop through thermal cycling of pipe, fittings, or flanges. Or, leaks will also show up where the air or nitrogen would not.




Given that there is a level of seepage from seals etc, how would you establish a de minimis leak rate method other than by using the working fluid in use after all the preparatory tests have been done?

.
________
mac game (http://macgame.org)

US Iceman
24-10-2005, 03:10 AM
I agree with that except that I’m not certain that I would want to introduce oxygen into a system containing air.

I have seen some preliminary reports that suggest the continued presence of oxygen in ammonia refrigeration systems promotes corrosion stress cracking in carbon steel vessels. For this reason (and efficiency) I always emphasize the need for non-condensable refrigerated purgers on any ammonia system.

I completely agree that dry nitrogen is the best.

Another potential area would also be the release of the nitrogen after the pressure test. During the blow down of the system my recommendation is that the release occur from a low point in the piping, system, or vessel.

My logic for this is to sweep any particle or debris out of the system. Any debris removed in this manner reduces that which would be removed by the strainers or oil filters.

My experience has been such that corrosion (oxidation) is very uncommon in ammonia piping or vessels. Any decommissioned pipe or vessel has always been relatively clean. My greater concern is the external corrosion and the reduced wall thickness that results.

If possible I would be interested to read about the accidents and the perceived cause.


Given that there is a level of seepage from seals etc, how would you establish a de minimis leak rate method other than by using the working fluid in use after all the preparatory tests have been done?

A very good question.

The pressure test is only as good as the final test pressure and the fluid used. In a similar fashion as helium leak testing, if the system under consideration will hold helium it is considered tight.

I have seen ammonia system pass a pressure test and/or vacuum test. Then, after several weeks or so in operation small amounts of ammonia can be detected.

Some of these are due to thermal cycling of the threaded joints through expansion and contraction. Other leaks have been found where the joint passed the pressure test (with ammonia present), but eventually leaked. Some of this I attribute to the joint compound that was used for the threaded joints.

Most of the seals will leak a few cc's of oil during a given time period. I consider the oil to be similar to a barrier fluid. The fluid leaks to prevent the refrigerant from escaping in larger volumes.

The final leak checking operation is after the system has been up and running for several weeks. It is a necessary problem to solve during the commissioning of the systems.

I also believe part of the problem is caused by the ammonia itself. Since ammonia is a very good solvent, it will clean all mill scale, rust, etc. from the internal surfaces of the system.

Threaded joints have seemed to be the most common issue. The thread cutting dies, threading practice, and joint compound seem to be a recurring issue.

After an initial shake down period, the systems can be operated with no noticeable trace of ammonia in the ambient air. This is a good sign that the system is well maintained.

This is simply an area which requires constant monitoring, and correction all through the life of the system.

Fugitive emissions of ammonia should never be tolerated.

Chopper
24-10-2005, 05:19 PM
My feeling about this method and what appears to be yours iceman is that this is not a good method of practice. I for one had it drummed into me from the first day i started as an apprentice and went through college that moisture in any fridge plant large or small, ammonia or R134a etc is a cardinal sin and is our enemy, so much so we do everything in our power to get rid of it. Yes i know that in large plant it can come in through thousands of joints, flanges, shaft seals etc, but at the end of the day it does not mean that we should over look good practice.
If the said people pressure test like this what else are the "cutting corners" on. Is it a full vac to start with?
Is this in compliance with the PED regs? Whats the point in getting every valve, vessel and the like all certified if it is put together and tested in such a manner. Is air denser than OFN?
At the end of the day this industry reguires a proper governing body to say what is acceptable and what is not, if not to clear up any confusion and raise standards!:)

MRcoolingMAJIC
24-10-2005, 05:37 PM
Some of these are due to thermal cycling of the threaded joints through expansion and contraction.

I'm wondering what the procedure was varifying this theory? Was there a control? Often times, when in doubt, thermal cycling is the default explanation.

US Iceman
24-10-2005, 06:19 PM
Often times, when in doubt, thermal cycling is the default explanation.

Granted this is a typical response. I have seen numerous threaded joints that seem to "loosen up" after the initial weeks of operation.

The only way I know to completely verify the cause is to pull the joint apart.

Was the joint loose when it was unscrewed? Check the threads with thread gauges to determine the proper thread depth and pitch. Are the threads undercut or over-cut?

If the above was acceptable then we can be looking for joint compound problems. Is the compound the correct type? Was it applied correctly? (There is a proper method) An incorrect joint compound will develop leaks by allowing the ammonia to cut through, or dissolve a path through the threaded joint.

Lastly, was the leak the result of thermal cycling? This is just a fancy name to say the joint was not sufficiently tightened in the first place, if, all of the above was done correctly.

rbratlett
24-10-2005, 06:35 PM
[QUOTE Lastly, was the leak the result of thermal cycling? This is just a fancy name to say the joint was not sufficiently tightened in the first place, if, all of the above was done correctly./QUOTE]

i've wondered how much more pressure cycling contributes to leakage compared to thermal cycling. using vfd's for condenser fan speed control having pid algorithms compared to fan cycling around a differential.

bet then again, what do i know?

US Iceman
24-10-2005, 06:40 PM
...at the end of the day it does not mean that we should over look good practice.
If the said people pressure test like this what else are they "cutting corners" on.

Makes you wonder doesn't it?

I think some of the differences come into play between definitions of codes/regulations, industry practice, and what people find "acceptable".

If you are working for the owner of a refrigeration system, you want the best you can possibly get for a reasonable price. In this case, would the owner have specifications setting out the minimum practices to be followed for his system?

After all, it is his system and he should have some say in how it is prepared and operated.

I'm sure there is a British safety standard for compliance of ammonia refrigeration systems. This sets the minimum requirements for construction and design.

Best practices are the methods we use to develop operating strategies, commissioning procedures, etc.

One thing I am sure of is that if the industry does not provide reasonable best practices that are based on logic and sound engineering principles, the government will ultimately force you to do so.