Outside CHW pipe rusts when system is drained during winter

[BronYrAur]

Looking for recommendations of how to best prevent pipe from rusting. I have a situation where some air-cooled chillers are drained during the winter for freeze protection. Upon re-filling in the Spring, considerable rust is showing up to the point of plugging fan coils.

The system is piped similar to a primary/secondary. There are primary pumps that feed the chiller, but the "secondary" pumps are actually dual temperature pumps that circulate hot water in the winter.

I think adding glycol to the whole system is out of the question. I don't know what effect that will have on the fan coils. Plus, this job is for three 20-story condos, so we are talking about a lot of glycol and mess during maintenance.

I thought about replacing the "decoupling" pipe with a heat exchanger and just putting glycol in the primary loop. I would have to see if the chillers have enough capacity, but do you think of that idea?

My only other idea is the charge the piping with nitrogen during the drain-down season. What do you think of that?

Thanks for the feedback.

 

[ChasBean1]

I agree about the nitrogen (or argon) charge. Just have a regulator off the N2 bottle to keep from overpressuizing your piping.

 

[BM4BRANES]

I too like the N2 purge idea, why are you apprehensive?

 

[BronYrAur]

I'm not - I just want to consider all options. It would be nice to not have the maintenance requirement of the N2 as they already have to put glycol in the chiller (because they can't get it completely drained). I thought a glycol solution would be less hassle. .. but more costly.

But let's pursue N2. Would I have to pull a vacuum before charging with N2? Otherwise, I would think that I won't get a pure N2 charge.

What pressure would you recommend?

Am I correct in my calculation that my volume of pipe (i.e. so many cubit feet) is equivalent to 1 atmosphere of air? In other words, if I want to charge the system to 2 atmospheres or 29.4 psia (14.7 psig), I need to have a cylinder of gas with 2 x the volume of my piping.

 

[jistre]

You don't necessarily need a pure N2 charge to retard corrosion. You just need enough nitrogen that concentrations of things like O2 and water drop low enough to retard corrosion. In the past, I've had vendors use two methods. One caveat, this all applied to new equipment, so all I really had in the equipment at the beginning was normal air.

One calls for blocking in the equipment, then raising the pressure to a given amount above atmospheric and holding it there for a time. This raises the N2 concentration inside the system above atmospheric concentrations and allows it to mix well. Then, you bleed off the nitrogen enriched gases inside the system to atmospheric pressure and repeat the process. From the initial N2 concentration, the purity of the N2 source, the pressure you're going to raise the system to, and the desired final N2 concentration, you can calculate the number of cycles required to reach that final concentration.

The other method I've had vendors use is to block the system in and hook up the nitrogen source and open a vent. Use a meter to measure the N2 or O2 concentration of the gas coming out of the vent, and once it the concentrations reach the level you want, then close the vent and after reaching the desired storage pressure, shut the N2 flow off.

In either case, you'd probably want to have a system in place so that after inerting, you can add N2 as needed to ensure the system is held above atmospheric pressure to prevent the system from being infiltrated by air with water and higher O2 concentrations.

 

[jistre]

Granted, my first method could be accomplished by pulling a vacuum, then relieving the vacuum with a pure nitrogen source, but vacuums do NASTY things to systems and vessels not designed to withstand them.

Something else I've heard of before, but have never used is to fill the system with liquid, hook a nitrogen source up, then drain the system. Basically, when you vent the system to allow it to drain, you introduce the high purity nitrogen as the vacuum relief gas, not even letting oxygen get into the system in the first place.

 

[jistre]

Check out AIChE's publication "Properly Purge and Inert Storage Vessels". It specifically applies to inerting systems to prevent explosions, but it should have some useful information and be better technically than my previous posts.

One other thing you'll want to consider when deciding on nitrogen inerting is whether or not the system you're dealing with could ever produce a low oxygen environment that people could enter, either through leakage, access to the system, etc. Depending on the concentrations, it doesn't take more than a few seconds of breathing a low-O atmosphere to pass out and less than a minute or two to die.

 

[BM4BRANES]

I would not spend any time on a vacuum, the volume to squeezed is enormus and coupled with the dehydration of any residual water, well, it would be summer if it got done at all.A simple complete purge with under no pressure ought to accomplish the task. You will need to incapacitate the air removal devices for the duration.

 

[BM4BRANES]

Bron, I've assumed there is a closed loop corrosion inhibitor (nitrite)when filled, yes? If the corrosion occurs during wet periods, you may be doing all the right things for all the wrong reasons.
BM

 

[BronYrAur]

BM,

You are correct. When filled, there are corrosion inhibitors.

 

[cjw81]

Eutectics (heat transfer) is part of what you need to address. I have had a lot of experience with cooling tower lines and rust and all due to emptying the lines of water in the winter season. Oxygen reacting with the piping will get you everytime. Once you see the strainers clogged with the insides of your pipes and your chiller going down due to a high delta T or delta P you need to get a solution quickly. A suggestion is to use a eutectic fluid like "Dowtherm" by Dow chemical. It is a heat transfer fluid that will protect your piping down to the temp you spec for protection.At the beginning of each season you send a sample of the fluid to Dow and they sell you only the inhibitors that were used in the season, you do not replace the entire medium.You could also consider, if fesable running the pumps when below 32F.Running water doesn't freeze.
You could also use proplyene glycol,relatively inexpensive.
Make sure to use an ooptical spectrometer for an accurate freeze protection temperature of the medium.

 

[BronYrAur]

I am liking jistre's idea of letting the nitrogen act as my vacuum relief gas. I could put a cylinder bank on the roof next to the chiller and connect to the manual air vents up there. That way, as I drain down on one of the lower floors, I will introduce only N2 into the system. Agree?

So back to my question of how much N2 do I need. Some will undoubtedly be lost in the draining process. I will have approximately 2,000 Gallons of empty pipe, which should be around 270 cubic feet. If I'm thinking correctly, that will give me atmospheric pressure (0 psig) in the pipe. If I double that amount, I will have 14.7 psig in the pipe, correct? What would be a good pressure setting? I'll assume that 1 whole volume will be lost in the drain-down process, so I'll add a volume to the recommended pressure setting when selecting nitrogen cylinders.

 

[BronYrAur]

Ok, now I'm flip-flopping on the vacuum relief idea. It will be too cumbersome for the end-user. I think I'm going to drain down using air and then start purging/filling with nitrogen from the low side.

Anyone have a good guess on how much Nitrogen I will need to purge the air? My volume of empty pipe is around 270 cubic feet. As I purge, of course, my nitrogen is mixing with the air. I seem to remember a differential equation along these lines from decades ago assume good mixing (which I probably don't have here).

Anyway, any recommendations of how much nitrogen will be necessary to purge and fill the pipe?

 

[marcoh]

Maybe have a talk to a water treatment specialist. A closed system would have little need for corrosion inhibitor as all oxygen quickly disappears in the water. Might need to slug dose some more inhibitor and circulate it through the system before draining it.

Nitrogen purging should be effective too and relatively cheap.