This article was written by longtime contributor and RSES CM Jeremy Smith. Thanks, Jeremy!

Nitrogen doesn’t absorb moisture like many techs think that it does, and I think that we, as technicians, need to reevaluate the reasons for the “triple evacuation” process.

OK. Hold on, now. Put down the pitchforks and torches, and give me a chance here.

Take a minute or two and read what I have to say, and at least I’ll have a head start if you still want to come after me. Fair enough?

First off, let’s start with what nitrogen is. Nitrogen is a chemical element with the atomic number 7. It's an odorless and unreactive gas that forms about 78% of the earth’s atmosphere.

Nitrogen is not truly inert; it is mostly inert, meaning that it is generally non-reactive, but it can and does have some chemically reactive properties. They don’t affect much of anything that we work on in the HVAC/R industry, so we can safely consider it “inert” for our purposes, but it isn’t exactly inert. If it were inert, ammonia (NH₃) would not be possible, as nitrogen would not be able to react with hydrogen. But I kind of digress from the point I’m trying to make here.

My point is about using nitrogen to “absorb” moisture during the evacuation and dehydration of a refrigeration system using a process commonly called “triple evacuation.” Quite a lot of techs seem to think that nitrogen has some special properties because it’s “dry” and will remove a bunch of water from a system. That’s not really how things work, and I’ll explain why that is if you’ll bear with me a bit. As we work through this together, I will try to minimize the number of confounding variables. That is both to keep everything consistent and for the sake of simplicity.

When we deal with nitrogen in our case, we will be referencing nitrogen at 70°F and standard atmospheric pressures. We are also going to assume that nitrogen is the primary agent in the air that absorbs or holds moisture and that no other gasses in the mixture that is air have any contribution to the ability of air to absorb moisture. That way, we can simplify the thinking and the math involved, and we will give nitrogen the benefit of the doubt. Increasing the pressure of nitrogen above atmospheric will reduce dramatically the amount of moisture the nitrogen is capable of holding, so we will keep it at atmospheric, again, to give nitrogen the benefit of the doubt.

First off, let’s look at what “dry” means in this context. What is “dry nitrogen?” Industrial grade “dry” nitrogen is 5 parts per million (ppm) moisture or lower. For convenience, we will use 5 ppm as a reference to maintain constancy and evaluate the worst possible case scenario. This 5 ppm gives us a dewpoint of -86°F or 0.0187% RH. Now, we have a definition of “dry” nitrogen.

Like air or any gas, for that matter, nitrogen can hold a certain amount of water based on its temperature. For nitrogen, you can find this relationship in the psychrometric charts for air. Let’s look into the psychrometric charts and see just how much moisture nitrogen can actually hold.

In this case, remember that we are giving nitrogen 100% of the credit for the ability of air to entrain and hold moisture, even though nitrogen only makes up 78% of air. At our selected conditions of 70°F and atmospheric pressure, one pound (by weight) of nitrogen will hold 100 grains of moisture at 90% RH. For reference, one grain is 1/7000th of a pound, so that 100 grains is 0.0143 POUNDS of water. An ounce is 437.5 grains, so one pound of nitrogen can hold less than a quarter ounce of moisture at 90% RH at 70°F.

How does that apply to the practice of triple evacuation and using nitrogen to try to dehydrate a refrigeration system?

Let’s start with a couple of more reference points. Then, we’ll get into the nitty-gritty.

At atmospheric conditions, one pound of nitrogen occupies 13.8 cubic feet. That means that a 40 CuFt cylinder of nitrogen that is completely full holds approximately 2.9 pounds of nitrogen.

I like to use hypothetical examples that closely resemble what we see in the field, so let’s set up a hypothetical example as a way to think about this so that we can see how difficult a process removing moisture with nitrogen alone would be.

In our example, we will consider a system that conveniently holds exactly 1 pound of nitrogen at atmospheric pressure and 70°F. Keep in mind that the internal volume of that system is 13.8 cubic feet, so this is a very large system. As a point of reference, 100 feet of 1 ⅛” ACR copper has an internal volume of 0.69 cubic feet, so we are using a very large system for this hypothetical example. Far larger than any residential system one will encounter. The results we get will be scaled down accordingly on smaller systems.

We will introduce one ounce of water into this system. For a little perspective, this is 2 tablespoons full of water. That is not very much water, given the size of the system.

What will it take to remove any significant amount of moisture using nitrogen?

The moisture, in our example, is simply a small pool or puddle of moisture, say, at the bottom of a trap or simply laying along the bottom of a horizontal section of piping, more than likely a victim of poor practices. This will be a new installation job, so we don’t have any oil in the piping that could cause a problem during the evacuation and removing any other confounding variables from the issue. We just have a puddle of water, and we need to get it out of the piping.

For the 1 pound of “dry” nitrogen to absorb the full 100 grains of water that it is capable of holding, it needs TIME. It needs to sit long enough for the humidity to diffuse throughout the entire system and raise the humidity of that entire quantity of nitrogen to 90%. If you only allow it to sit long enough to raise it to 30% or 40%, for example, you only pick up between 30 and 40 grains of moisture, which is only a fraction of the total maximum that same pound of nitrogen can hold.

In a system of this size and complexity, the time to diffuse the humidity through the entirety of the system could be several hours or longer. Then, you have to repeat the standing nitrogen step once more, again waiting several hours (or longer) for the humidity to diffuse throughout the system again. We do all that for a measly 100 grains of moisture, less than a quarter of an ounce.

Remember the basics here. Everything moves from “high” to “low,” so the moisture moves from an area of high humidity to an area of low humidity. This process is rather slow. Imagine, for purposes of scale, what effect a small pot of water would have on the humidity of a room if that water were left at room temperature. How long would it take for the higher humidity level close to the water’s surface to have an impact on the humidity 20 feet away? Even if we lowered the humidity in the room to the levels of “dry” nitrogen, the humidity diffusion rate would be very low and very slow. Would it ever have a measurable, appreciable effect on the humidity of that room? This is key to why nitrogen doesn’t really have any impact on evacuation procedures.

While evacuating the system, there is a constant removal of moisture as the water will constantly be boiling, constantly absorbing heat from the surrounding piping and materials, resulting in a faster moisture removal because there are no interruptions in the process.

Well, wait, you say. I’m not saying that nitrogen “absorbs” water, but nitrogen DISPLACES moisture. OK, to a limited degree, this is a fair statement. IF you have a system filled with warm, humid air, a gentle push or ‘flush’ of nitrogen will displace that moisture. I’m not entirely certain how this displacement is expected to accelerate dehydration significantly. It will displace moisture one time. A vacuum pump will do exactly the same thing and do it just as fast.

Remember, the moisture content of air is measured in grains of moisture. The actual quantity of moisture is extremely small.

But wait, I’m not using nitrogen to “absorb” moisture, and I’m not using it to “displace” moisture; I’m using it to prevent the moisture from freezing because the vacuum is being pulled so fast that I don’t want that to happen. If this is your position on the use of nitrogen, I’d first ask you to listen to the HVAC School podcast episode with Jim Bergmann, where he specifically addresses this issue with a resounding NO.

In a normal HVAC/R system, under conditions that are not near or below freezing, it is extremely unlikely to freeze moisture due to the amount of heat available in the environment that will keep any liquid water in liquid form and keep that liquid water boiling under vacuum. This simply is a non-issue.

But wait, Jeremy, you say. I KNOW nitrogen helps speed up my evacuation because I use it all the time, and it makes my evacuation go faster. I’m not exactly sure what to say in cases like this. I think I’ve made a pretty convincing case that nitrogen doesn’t really assist the evacuation process. The math and the science are really on my side here. In many cases, particularly in the service world where systems are ‘contaminated’ with refrigerants and other gases, nitrogen can just kind of recalibrate the micron gauge and cause it to show you a true reading rather than one that is distorted by a non-nitrogen based environment.

The last commonly used reason for triple evacuation is “the book says so” Yes, it generally does. If your installation manual, your customer, your employer, your foreman, manager, or anyone else that has authority over the job you are doing or your job, in general, says to do a triple evacuation or any other process or procedure that you deem useless, DO IT unless that process is unsafe for your person. In no way, shape, or form am I encouraging technicians to refuse to do work as they are expected to do it because some random dude on the internet said so.

To conclude, yes, nitrogen does absorb some moisture. It really isn’t as much as most people like to think it is when you break it down and look at the math. It also takes a lot more time to have an appreciable impact than most techs are willing to give the process for what little benefit it can offer.

So, what is the takeaway here? Why did I spend a bunch of time working on this, researching the math, and laying out these ideas in a way that is (hopefully) easy to read and digest? Do I honestly expect manufacturers to change their processes and procedures based on my couple of pages? Nope. I don’t. If anything, my goal here can be stated in a single word:

THINK!

Think about what you are doing. Think about why you are doing it. Evaluate every process, technique, procedure, and idea that is given to you. Don’t take something at face value just because you trust the person telling it to you or because you like him or he’s the boss. I’m not trying to organize some industry-wide revolution, so don’t tell your boss that I said he was wrong for making you do a triple evac. Don’t tell the Daikin or Mitsubishi rep on a VRF startup that I said that triple evac was stupid and that you shouldn’t do it. I haven’t said these things. I hope you can take away a little thoughtfulness about the deeper reasons you’re doing the things you’re doing.

—Jeremy

P.S. – Bryan here. I was talking to Eric Mele about the subject of purging with nitrogen and triple evac. While it's pretty clear that purging with nitrogen or breaking with nitrogen is going to do very little to help with liquid water contamination, it does help you get a more accurate reading on your vacuum gauge (as Jeremy stated). It has one other impact, which is that it ensures that a larger quantity of molecules leftover in the system will be “dry” nitrogen rather than oxygen or water vapor (which are unwanted) simply through “homogenous dispersion.” That simply means that the more times you run nitrogen through the system, the greater the concentration of nitrogen molecules vs. other stuff even under deep vacuum.

The undisputed conclusion is that proper assembly practices and deep vacuum are keys to keeping the nasty stuff out of a running system.