< Back to All News Nitrogen Generator for Fire Protection and Fire Sprinkler Monitoring System

Automatic Air Vents to Control Oxygen Corrosion

This presentation about automatic air vents and advanced corrosion solutions was delivered at the NFPA Conference and Expo in Las Vegas to address the 2016 Edition of NFPA 13 that requires a means of venting air on all wet pipe sprinkler systems. 

Benefits of Automatic Air Vents to Control Oxygen Corrosion in Wet Pipe Sprinkler Systems

Video Transcript:

We’re going to talk about the benefits of automatic air release valves on wet-pipe fire sprinkler systems. But before I get into why we are where we are now, I want to go back to how we got here.


In about 2000, there’s a movement that started in the late nineties regarding MIC, Micro-biologically Influenced Corrosion. When I got into this business in 2008, every fitter on the planet knew what MIC was. MIC was causing leaks in fire sprinkler systems. And MIC is a real thing. Organisms produce conditions under which metal fails. They dissolve the metal. A lot of them produce acids. But MIC was a big thing, and it became a bigger thing when in 1999, NFPA 13 required that people start paying attention to MIC. The code language from ’99 edition suggested that, “In areas where water supplies are known to have contributed to MIC, water supplies shall be tested.” Now they didn’t say how to test it, they just said they shall be tested. If you talk to any microbiologist, how you test it is very important. And they said, “Appropriately treated,” but they didn’t say what that was either. But we’re well on the way to MIC, MIC, MIC.

This started gaining momentum when FM Global produced their data sheet 2-1 in 2001, Prevention and Control of Internal Corrosion in Automatic Sprinkler Systems. And if you look at the term MIC, it shows up all over this data sheet. In fact, the word MIC is used 55 times in this data sheet. So it was all about MIC. So within the context of it they talked about enumerating sulfate reducing bacteria and iron reducing bacteria, using disinfectants. This got into the code for the US government, the DOD, the GSA, everybody was talking about using chlorine and microbicide to control MIC, because the understanding was MIC was the cause of corrosion.

So in this, what I call the MIC, MIC, MIC era, the fire protection industry was totally focused on MIC as the primary cause for leaks. Industry experts, FM Global, NFPA 13, all recommended management of MIC. So that’s where we were. As a result of that focus, there was widespread use of microbicide, cleaning agents, corrosion inhibitors, and the pipe manufacturers started making MIC coatings. And those coatings were designed to control corrosion caused by bacteria. NFPA 25 even began requiring 5-year mic deposit tests. Again, there was a lot of emphasis on pulling samples and measuring for bacteria, but there wasn’t a clear understanding about what to do about it.

ECS History

My company’s been doing MIC deposit tests for 10 years, the code requires it. What we found out is there are not sprinkler systems in the country, I’ll say in the world, that doesn’t have lots of bacteria in it. They’re filled with bacteria. Primarily because these are nice, warm, stagnant, cozy conditions. You have residual cutting oils from threading machines. You have debris from the pipe itself. You have all kinds of nutrient materials that organisms can live and grow off of. So they’re found in every system. We’ve never pulled a sample from a system and not found bacteria in it.

But we also found that there’s no correlation between the quantity of bacteria and the number of leaks. So you could have a lot of bacteria in the samples from your system and no leaks, or you could have very few bacteria and lots of leaks. There was zero correlation. What we saw however, was there was a direct correlation between the number of leaks and the number of drain and fill cycles. Malls, for example, where they do lots of tenant modifications on the sprinkler systems, and they drain and fill and drain and fill a lot, had a lot more leaks than office buildings where they never drained and filled at all.

So we started looking at other causes. We did a lot of failure analysis. We do roughly four to five hundred pipe samples a year, and all of the metal loss patterns we saw indicated that the corrosion was caused by oxygen. Simple oxygen corrosion. We had done a lot of work in the past on analyzing MIC corrosion. MIC has very characteristic metal loss, it’s concentric stair-step plateaus. And the reason they are caused is because an organism attaches to a pipe, it goes from a planktonic state to a sessile state, consumes nutrients from the water. And as the colony grows, you get metal loss under the colony itself. We didn’t see this very often.

We said 7 or 8 years ago that the failures attributable to bacteria were less than 10%. We suggested less than 5%, but everybody was still out there talking about MIC. The most common leak we see in fire sprinkler systems is weld seam corrosion. The reason that is is when fire sprinkler piping is welded, the heat affected zone around the weld becomes more prone to give up its electrons and have corrosion. It’s a very well known phenomenon. A lot of industries that use electric resistance welded pipe actually heat treat the pipe after it’s made to eliminate that heat affected zone. But this industry couldn’t do it because of weldoletes on the manifolds. You couldn’t heat treat that pipe. So we started looking at other options for controlling the corrosion.

Another thing we found, we were the first to point out galvanized pipe fails three to four times faster under identical conditions to black steel because of the oxygen corrosion. There’s lots of complications with galvanized pipe. We’ve seen failures in Las Vegas in twelve months. When you have trapped water in a main and a weld seam under that trapped water, you can actually get through the wall penetrations in less than a year. And it’s because of the unique characteristics of the zinc coating and how it corrodes. So we began recommending 7 or 8 years ago, “Stop using galvanized pipe.”

There’s other issues around it. We published a paper called Six Reasons Why You Shouldn’t Use Galvanized Pipe in Dry and Pre-Action Systems. One of them is very simple. The zinc that is shed during the corrosion process on galvanized pipe is a toxic heavy metal. And what do we do with full flow trip tests after we get finished testing a system? We dump that to the sewer. Well, most cities, this city probably has a limit to discharge of less than 50 milligrams per liter, most samples that are going out have between 1000 and 2000. So when you do that, you’re committing a felony because you’re discharging heavy metals, but nobody knows about it so nobody worries about it.

The bottom line is, galvanized pipe fails and FM Global has now recognized the acute vulnerability of galvanized pipe. In their last few publications, they’ve indicated galvanized pipe can fail in two to three years in dry and pre-action systems where there’s trapped water. So it’s a big problem. FM says MIC is responsible for less than 10%. We think it’s now less than about 5%. MIC occurs. MIC is a problem. It’s a real thing, but it doesn’t cause many leaks.

So in 2009, we published an article in Sprinkler Age magazine entitled MIC is Not the Primary Cause of Corrosion in Fire Sprinkler Systems. We said it’s oxygen. So let’s start thinking about oxygen. When you look at failures in fire sprinkler systems, when they’re fully replaced, only about 80% of the pipe shows any corrosion damage at all. In wet-pipe systems, the corrosion is localized to where the trapped air is. And in dry-pipe systems, the corrosion is localized in the mains where there’s trapped pools of water. So it’s very, very localized.

In order to have corrosion with oxygen, as a corrosion engineer, we focus on three kinds of corrosion gases: hydrogen sulfide, carbon dioxide, and oxygen. Oxygen is a very damaging, corrosive gas. In order to have oxygen corrosion, you must have liquid water because the oxygen dissolves in the water and then reacts with the metal. If you don’t have liquid water, you cannot have corrosion. So you won’t have corrosion in freezers, for example. The temperature slows down the reaction rate, but no liquid water, no corrosion. If it can’t dissolve in the water, it won’t cause the chemical reaction.

So here’s two things you need to understand about the physics and the chemistry of oxygen corrosion to get how we came at our solution. It’s a two step process. Oxygen dissolves in the water first, so it has to dissolve in the water. Well, that’s a bottleneck because oxygen solubility in water is very low. You can only get 10 parts per million of oxygen to dissolve in the water. And when I say dissolve in the water, it has nothing to do with H2O. Those are covalent bonds, and the O in the H2O is not involved in corrosion at all. It’s the oxygen that fish breathe. So 10 parts per million is .001%, very tiny amount.

Soon as the saturation level hits 10 parts per million, no more oxygen goes into the water. And remember, in order to have corrosion, it has to dissolve in the water first. So it has very limited solubility, but it also , in stagnant systems, has very limited mobility. That means the surface where the oxygen in the air contacts the water, it’ll dissolve into the water, but it’s not going very far because the oxygen molecules are not moving very far from the point where they enter. So you get it to dissolve in, it hits saturation, that’s a bottleneck.

So a trapped reservoir of air in a fire sprinkler system takes about 90 to 120 days to get consumed. Why? Because it dissolves into the water, it reacts with the pipe, falls below saturation limit in the water, more dissolves in, and you get that progression until it’s consumed. In 120 days, it’s all gone. Okay?

Wet System Configurations

So if we look at a typical gridded wet-pipe system that’s pitched in the center of a building, You have trapped air. It’s very important to understand, in order to have corrosion you must have liquid water, air, and metal. Can you have corrosion in that point where the trapped air bubble is? No. There’s no water. There’s no water for the oxygen to dissolve into. Down here, where it’s fluid packed, you can’t get corrosion either. Why? Because it’s going to dissolve into the water at the air-water interface, and remember I said low mobility? It’s not going to move very far down here. So you’ll have very little corrosion here. All of your corrosion is going to be in close proximity to the air-water interface, okay? That’s very important. That’s why you have such limited pipe damage. It’s damage where the air-water interface is, and you get secondary corrosion where the debris ends up, but it’s not corroded at all up at the peak or where it’s fluid packed.

Here’s another typical configuration. Front of a building to back of a building. This is a Walmart, Home Depot. All of those stores are pitched from back to front. So all the trapped air is at the front of the store. Why? Because that’s where the cash registers are, and that’s where we want our leaks to occur. Seriously. You know, you can go on … Nevermind

So all your leaks are occurring at that point where the air-water interface is. So two measurable things occur as a result of this corrosion reaction. One is, it creates a pit. That metal that reacts with the oxygen leaves the surface of the pipe and becomes a water-soluble ion. For a very short period of time it’s water-soluble, it reacts with the oxygen, and it forms iron oxide, which is the second component, sludge. So you get iron oxide rust, which is incredibly insoluble. If you ever got a rust stain on your front porch and you try to get it off, it does not want to dissolve.

So here are some pictures, air-water interface, main, branch line. This is a high branch line, look at all the corrosion near that interface. Look at the corrosion near that interface. So, here’s an example. This is a client in California who had a 400,000 square foot warehouse, and he started having leaks. And this configuration here’s the main, here’s the riser nipple. And this is the end of the branch line, and it’s pitched up. So where’s all the air going to be? Right in the end of that branch line. So when he started having leaks in the building next to this, he had enough leaks, he had enough adverse effect on his business operation … They actually stored energy drinks. The fire sprinkler contractor tore the whole system out and replaced it, 400,000 square feet.

On this one we said slow down. We’re going to show you that the corrosion is very limited. It is not everywhere. Because if you go talk to people who have a warehouse and you say, “Where are all your leaks?” You know what they say? Everywhere. Because in their mind it’s occurring … because it’s driving them nuts. But the reality is that when you analyze it, it’s occurring just where the air-water interface is.

So here’s what we’re going to look at. We’re going to open up a piece of pipe. So you know what the fitters say the first time they see this? MIC! You got MIC! And the reality is, if I pull a sample of this and culture it, there are millions and millions of organisms in there. But the most common organism we find metabolizes iron oxide. They actually use iron oxide as a nutrient source, okay? So what we now know are the bacteria are the result of oxygen corrosion, and they did not cause the corrosion. Does that make sense? It was misdiagnosed, and we went down the wrong road for a long time.

Venting air on all wet pipe sprinkler systems

So our solution was get rid of the oxygen. I’m going to show you a video … This is a high branch line, right at the air-water interface. And just look at the tubercles. If you pop these tubercles off you can see the air-water interface. There’s going to be a pit underneath each of the tubercles, and when the pit gets deep enough, you’re going to get a leak. And from my perspective, and from the fire sprinkler contractor’s perspective, they know with 100% confidence, when I get leak number one, I am going to get leak number two. Why? Because the process of repairing leak number one, what do they do? Drain the system down, put a bunch of new oxygen in there, fill it back up. I’ve just recharged that corrosion mechanism, and it gets worse and worse over time, so it accelerates. You actually see leak frequency increasing.

So, another example of how you only have corrosion localized at the air-water interface, in our video scoping we can tell how far we are in from the entry point. So this is 37 feet in from the entry point going up toward the pitch, and you can see lots of corrosion. The next slide is 4 feet in, where it’s fluid packed, where there’s no air-water interface. You have very little corrosion damage. So we do a lot of systems, we video, this is the pattern we see. You get corrosion at the air-water interface, but you don’t get very much corrosion at all where there’s all air or where there’s all water.

So everybody has seen, why is this water so brown and nasty? It isn’t MIC. That coloration you see is iron oxide that’s been stripped off the pipe further up in the system at the air-water interface. See all these rust stains? I worked with a contractor with Western States Fire Protection in Texas, and we were having difficulty figuring out where the riser room was, and he said, “It’s easy. You just go in the parking lot and you look for the big rust stain.” Why? Because the main drain dumps out onto the parking lot, and all of that rust is from your pipe being dissolved by oxygen up in the system. And it goes on long enough, you’re going to end up having leaks.

Once oxygen and trapped air was recognized as the root cause, venting became a good solution. And here’s something critical you have to realize: People get all worked up about “I got to have a vent. NFPA 13 says I have to have a vent now. Where do I put it?” In our opinion, it doesn’t matter where you put it. But I wouldn’t put it up high on one branch line because it’s only going to vent the air from that one branch line. I’d rather have it on a riser nipple on a far main because remember when they used to have inspectors test at the end of the zone? As you were filling it up, you’d have an apprentice out there, and he’d be told, “Go open that valve. And when you see water come out, close it.” We vented a lot of air from these systems during the filling operation. As we evolved, we moved the inspector’s test to the riser. None of that air gets vented anymore. So it’s all stuck in the system.

As a chemist, I’ll tell you. This corrosion reaction is linear. That means if I vent 25% of the oxygen, the cumulative corrosion that occurs in that system will be reduced 25%. If I vent 50% of the oxygen, 50% of the corrosion will not occur. Why? Because that oxygen molecule isn’t stuck in the system anymore. So put a vent on! It’s the cheapest thing you can do to control corrosion. Now we take it a step further with nitrogen, but putting a vent on is the most simple, inexpensive means for reducing corrosion.

Here’s the language in NFPA 13 2016 edition. People fought this tooth and nail. For years they didn’t want to put a vent on. But they finally came around to the benefits of having air removed with an automatic air vent. Now you can put a half-inch ball valve up there, but what is the likelihood that you’re going to get somebody up on a lift to go open that ball valve every time you fill it? I’ll tell you what the likelihood is. Slim and none. So putting an automatic vent is a very inexpensive way to make sure that you’re venting gas every time it fills.

So in the latest update of the FM Global 2-1, they not only discuss the use of nitrogen, they also almost eliminated the discussion of MIC, but they included language regarding the root cause, water and oxygen. remove trapped air from wet-pipe systems, install FM approved automatic air release, or manual, remove the air each time the system is drained and refilled. As a chemist, I’ll tell you, I know with mathematical certainty every molecule of oxygen you trap in a wet-pipe system will react with the pipe to completion and be consumed within 90 to 120 days. And you will remove a fixed amount of metal from the pipe.

So here we are back to our trapped air spot. So remember I told you, this graphically doesn’t depict it. These air-water interfaces can be twenty feet long. Why? Because if you have a mild pitch, it can be twenty feet long. But that’s where all the corrosion is. So here’s something else very practical. If I vent some of that air while I’m filling this pipe, what’s going to happen to that air-water interface? So I put a vent on a far main, and I vented 50% of the air that would have otherwise been trapped.

What’s going to happen to that air-water interface? It’s going to move. Where’s it going? It’s going up. Why? Because I have less air trapped. But what do I know about this metal up here? It’s not corroded like this metal. Nor is it corroded down here either. So the cool thing from a physical, chemical standpoint, if I vent 50% of the air, I’m moving my interface to new pipe. I’m going to make that damn thing last longer just because I vented air. So don’t get all worked up about putting a vent high on the branch line. Just put a vent on a riser nipple where you can get  50 to 60 to 70% of the air out, because it’ll happen. And then you move the air-water interface. Then a piece of pipe that would otherwise have failed is going to be working on new pipe. Does that make sense? So the benefits are very, very significant of just putting a vent on. Any kind of vent.

ECS Ejector Automatic Air Vent – PAV-W

So the vent we came up with is a really simple mechanical vent. I’m a simple guy. I like simple solutions. This is a redundant float valve. The components of it … It has a bushing, a ball valve, a union. Then we have a Y-strainer to protect the orifice in the first float. Here comes the gas, it goes from the discharge of the first float to the inlet of the second float. Mechanically, very simple. That redundant mechanical system is not going to leak. Why? Because it would have to fail in a lot of different aspects. We have a Y-strainer to protect that orifice.

So on this you see a high visibility gauge. That gauge is in the line between the first float and the second float. If the first float is functioning, the gauge will read zero because it cannot see the system pressure. That float has stopped the water from coming out, and you have zero. If the first float fails, you’ll see the pressure of the system on the gauge between the two floats.

We’ve sold thousands of these, and they don’t leak. They don’t have to be plumbed to drain. No requirement to plumb to drain. They’re the lightest vents in the industry. Here’s one I brought with me. So you don’t need a hanger. They have the lowest clear height, so you can get them into tight places much easier than the older vents that were out there. You can see this from the floor. So if this needle goes from zero into the yellow, you know that the first float has been breached and the second float is holding the pressure.

The assembly is FM approved for use on all wet-pipe fire sprinkler systems. A comparison of the others that are out in the industry, as I said earlier, it’s patented. We include the ball valve so it’s easy to just get it and put it in, low clear height, redundancy. We think that’s special because it virtually eliminates the risk of a leak. There’s no requirement to plumb to drain, no hanger required. You can see it from the floor. Simple connection. One mechanical T to put the valve in place. So in summary, remember: Reduction in the total amount of corrosion is directly proportional to the amount of air you vent. So I don’t care if you vent 30% of the air, you’re going to move that air-water interface and you’re going to reduce the amount of corrosion that takes place in the system by 30%. Our vent is approved, it works automatically, guaranteed not to leak, no hanger.

The last thing I put on here allows for future use with nitrogen inerting. So if someone asks, “What is nitrogen inerting?” So in a wet-pipe system, corrosion occurs wherever you trap oxygen. Air contains 21% oxygen. The best way to completely control corrosion is to remove oxygen from the system completely. So about 9 years ago, we worked with an engineer from Hughes and Associates and designed a system that removed all the oxygen from the piping before we filled it with water. We did three purges of the system with nitrogen to take it from 21% oxygen to basically zero. Once we saw that the discharged gas was nitrogen completely, we filled the system up with water. It’s a process that takes about two hours per zone. Once you’ve inserted the gas and converted it from air with oxygen to nitrogen, any trapped gas in the pipe is inert. You’re not going to have any corrosion.

We modified this air vent with a couple of fittings for sampling, and for preventing the nitrogen from getting out, and it’s a very simple conversion process. So with clients that want to stop corrosion, the first, least expensive move is to put a vent on. The second iteration in that whole design is to get all the oxygen out, then you have no corrosion. We have systems that were greatly damaged. We did inerting on them, the corrosion stops. Even with a really damaged pipe, it doesn’t have leaks anymore.

You can do it with any source of nitrogen. We used cylinders initially on the initial test, but we’re doing installations now. We just did a mall that had eight riser rooms, 53 zones, 1.6 million square feet, schedule seven branch lines. So you can imagine. They had 66 leaks in the campus on that one building the year before we inerted it. They were a very astute property owner. They said we believe nitrogen inerting will work. We put a generator in and we ran a one and a quarter inch rolled groove black steel delivery line with a quick connect manifold in each of the eight riser rooms so they had access to nitrogen any time they needed it. They did 75 tenant modifications per year, so they needed instant access. It worked. We inerted it about 18 months ago with no pipe replacement, and they’ve had three leaks in the 18 months, and all the leaks were associated with areas where they did some modifications.

So inerting is profoundly successful, and it points to and validates the concept that oxygen is the root cause of corrosion in these systems. If you get rid of the oxygen, you’re going to eliminate all of your issues associated with corrosion. But venting, putting an automatic vent on, is the first, simplest, least expensive route to pursue.


Just the numbers.