Freedom Spill - In a Perfect World What Should Have Happened, Part II

This is the second in a short series of posts about what probably should have happened in Charleston, WV a little over a week ago now when a spill of Crude MCHM shutdown the water supply of well over 100,000 people for most of a week. The earlier post (listed below) I described how the water company could have kept the water out of its system in a perfect world.

Freedom Spill - In a Perfect World What Should Have Happened, Part I

Of course, in a perfect world, the material never would have gotten into the Elk River in the first place. Let’s take a quick at what a perfect chemical company would have done to ensure that the product never would have made it into the river.

Materials of Construction

The first thing that anyone should look at when deciding to do bulk storage of any chemical is to determine what the proper materials of construction are to be used for the tank walls, the piping, the gaskets and various seals; all wetted surfaces. You don’t want materials that will react with the chemical or will be corroded by the chemical (actually the same thing, just with a slightly different focus). If the chemical acts as a corrosive to the material that is used to construct the tank then sooner or later there will be a hole in the tank.

The easiest way to determine the proper materials of construction would be to look it up in one of the standard industry references. Many common chemicals have listings of compatible materials on-line. For less common materials it may be as easy as contacting the manufacturer, particularly the larger ones with good engineering departments. The tried and true method, however, is coupon testing. You take a sample of the chemical and a sample (a coupon) of the material you want to use to make your storage tank. You put the coupon in the sample for a lengthy period of time. When the time is up you pull the coupon out and closely (with a microscope) examine the coupon for signs of pitting or corrosion. If you don’t find any such signs, you are good to go. If you do, try something else.

Freedom Industries bought their Elk River terminal with the tanks already in place. I don’t expect that you can find Crude MCHM in any of the on-line sources and their supplier (I’m not sure who their supplier is) may not have been able to tell them. With already existing tanks on site, I would suspect that there was no coupon testing done. Someone made their best educated guess and said that the Crude MCHM doesn’t look like it would be too corrosive, or maybe they didn’t even think about materials of construction. That will be one of the things that the CSB will be looking for in their investigation.

Preventive Maintenance Program

Even if you have done your job well in selecting your materials of construction, you still have to have a preventive maintenance (PM) program in place for any storage tank system. You have to have a routine set of inspections done to make sure that the tank and all of its attached piping are working the way they were designed.

One critical part of that PM program is the visual inspection of the interior walls of the tank for signs of corrosion or weakening. This means that the tank gets cleaned on a regular schedule (every 3 to 5 years, depending on what you are storing) and a trained inspector goes into the tank with a big flashlight and looks for signs of damage to the metal and the welds or other joints. Any signs of metal fatigue or corrosion call for a serious re-examination of the use of that tank for the storage of that chemical. Serious or significant signs of corrosion could cause the tank to be taken out of chemical service until it is repaired or replaced.

Another part of the PM program is less rigorous but just as important. Every time that employees work around a storage tank they should be trained to look at it and report anything out of the ordinary. In every leaking tank that I have ever observed, the leak always started out small. If a tank has to leak, you want to catch the leak before there is a puddle of the chemical on the ground; every pound on the ground is a pound that you can’t sell to your customers.

I have no idea when the last time that the Crude MCHM tank was tested. You can bet that the CSB will ask and ask for documentation. From the news reports and the size of the Crude MCHM spill I would be very surprised if anyone was reporting small leaks at Freedom Industries. The CSB inspectors doing their initial walk around of the site will certainly have been looking for signs of leaks on other tanks at the facility.

Containment Maintenance

As I have mentioned on a number of occasions any time you have an above ground tank farm, you have to have some way to keep potential spills from those tanks confined. Typically this means that an impermeable wall of some sort is constructed around the tank farm. The construction of this containment is not as simple as just putting up a brick or concrete wall. Where the walls come together with the floor of the tank farm (which should be an impermeable barrier in its own right to keep nasty stuff out of the ground water) provisions have to be made to seal the joint so that nothing leaks out.

Materials of construction are not nearly so important in a containment wall. You don’t expect the material to be left in contact with the wall for any long period of time. All spills should be promptly cleaned up so there shouldn’t be long enough contact with the wall/floor to cause a breach.

Preventive maintenance checks of containment dikes and walls are usually pretty simple and do not require extensive training. Most tank farms are open to the weather and fill up with water every time that it rains. The water should remain in the containment area until specific action is taken to remove it. If the maintenance folks are not complaining about having to drain the containment area after a heavy rain, you have a leaky containment area. The leaks are easy to find; just look for the small river running away from the wall. Even after the rain you’ll be able to find the dried creek bed.

Spill Control

Most spills should never make it to the containment wall much less into the nearby river. Again the vast majority of spills start out small. Facilities are required to have a spill control plan in place to prevent those small spills from becoming a big deal. Once the source of the spill and the hazards associated with it are identified employees should begin controlling the spill with a variety of tools and devices. The idea is to keep the spill confined to the smallest area possible so that cleanup is easier.

The other thing that is done quickly is to take steps to stop the flow of the material out of the storage tank. Again there are a variety of tools available to do this including patches and plugs that can even deal with relatively sizeable holes in the walls of tanks. I have personally seen a six inch hole in a tank plugged well before 5,000 gallons of product leaked from the tank.

The news reports that I have seen seem to indicate that there were no significant spill control efforts taken inside or outside of the containment wall until after the West Virginia Department of Environmental Protection (WVDEP) inspectors had actually gotten on site. At that point a Crude MCHM slick could already be seen on the Elk River.

Spill Detection

There are lots of ways to detect spills. For really dangerous toxic chemicals there are gas detectors that can detect the airborne component of the spill down to the parts per billion detection limit. For flammable chemicals there are detectors that can determine when there is a concentration of the chemical in the air that could be expected to explosively ignite. All of these detection schemes are expensive and only justifiable if there is some special risk for relatively small leaks.

Again, the most common way that leaks are detected in tank farms is individual employees seeing the evidence of the leak in the normal course of their duties. In this case, however, it seems that the odor of the product allowed WVDEP inspectors to find the leak based upon public odor complaints. It would seem unreasonable to assume that the folks that worked at the facility were not able to determine that there was a leak based upon the odor in the plant area.

A possible reason that the odor did not raise flags at the facility a week ago yesterday may have been that they had been living with the odor of a leak for some time. The human nervous system does a real good job of ignoring odors after a certain amount of time being exposed to the odor. This will be another item that the CSB will address in their investigation.

Next Installment

In the next installment in this series I will look at the communications issues associated with this spill and look at how they would have gone in a perfect world.

The Freedom Spill – Why?

While the folks directly affected by the spill are worried about where they are going to get their clothes washed and take a bath, the bigger question is why did this happen? Why did an industrial chemical get into a city water distribution system? The Chemical Safety Board is scheduled to have a team arrive tomorrow to start their investigation.

The investigation will probably take much less time than normal; investigating a leak where everything is still together is a lot easier than reconstructing explosions and fires. If there isn’t another conflict with DOJ on crime scene vs chemical incident scene investigation we should see something in a couple of months.

But, given the news reports that I’ve seen, access to Google Maps and a little knowledge of process chemistry I think I can piece together a pretty fair picture of what probably happened.

The Facility

Looking at the facility on Google Maps, it is clear that it is (as news reports have noted) that this is a chemical storage facility, not a chemical manufacturing facility. There are 10 large storage tanks clearly visible within the facility, three smaller tanks and a number of much smaller storage containers. The 13 identifiable tanks all appear to be within a brick wall that probably serves as the primary spill containment. The area encircled by the brick wall is certainly large enough to contain spills of all of the tanks at the facility.

It is also obvious that the facility is very close to the Elk River with not much more than a chain link fence separating it from the river’s high water mark. There appears to be a barge loading/unloading facility at the water’s edge. The rail line that can be seen along the eastern boundary of the facility is no longer connected to anything beyond the property.

There is a tank truck loading facility between the large tanks and the perimeter fence. Tank trucks come in the north gate at the facility, get loaded and depart through the south gate near the administrative building.

The Spill

At some time on Thursday morning a tank containing Crude MCHM started leaking. From news reports the leaking tank was a 48,000 gallon tank so it was one of the ten large tanks. Most facilities do not have leak detection equipment for their bulk storage tanks unless there is potential for catastrophic consequences and the Crude MCHM certainly does not fit into that category.

Most leak detection regimes rely on preventive maintenance programs to prevent large leaks and small leaks are typically detected by employees working in the vicinity of the storage tanks. With the Freedom Industries leak being as much as 5,000 gallons (about what an over-the-road chemical tank truck carries) this was a fairly large leak. It will be interesting to see what the CSB says about the source of the leak.

Containment Failure

The news report sited above says that the containment overflowed. That does not seem to be possible. The three foot brick wall around the tank farm should have been more than high enough to contain the entire contents of one of the tanks, much less 5,000 gallons.

Ken Ward reported that the West Virginia Department of Environmental Protection (WVDEP) has ordered Freedom to empty all of their tanks at the site because the containment was not adequate to hold a spill. I suspect that the brick wall has cracks in it near the foundation that would allow any pooling liquid to leak out. This would have been obvious to the facility maintenance team every time that it rained and water collected within the containment area.

No Spill Response

The third part of any spill containment plan also failed in this instance. Any time a major leak is detected in a tank farm, one of the things that has to be done immediately is to check to ensure that the spill is being contained by the containment barrier. Any leaks in that barrier need to be controlled with mobile spill control equipment. News reports indicate that the WVDEP inspectors saw no such efforts being made when they showed up at the site.

The Water Company Response

News reports from WSAZ indicated during the initial response on Thursday that the water company felt that their inlet treatment system would deal with the relatively low amount of contamination (I have seen figures of 41 ppm for this spill) in the river water used to feed the water treatment plant. This is certainly true for a wide variety of contaminants that such facilities typically see.

This is not a chemical that the water company is required to test for, so there was no method of identifying this chemical in their water readily available on site. There are literally tens of thousands of industrial chemicals and testing for their presence in water can be quite difficult for a well-equipped water testing laboratory. Very few water treatment facilities can afford such facilities and most rely on out-side labs for non-routine testing. On site labs only test for those chemicals identified by State and federal environmental regulations; Crude MCHM does not show up on those lists.

According to news stories water system samples were sent to an outside lab for analysis even though the treatment facility thought that their system would remove the chemical upon intake. When that proved not to be the case, the emergency ‘no contact’ order was issued even though there is no clear indication that there was (or was not for that matter) any hazard associated with the level of contamination in the system.

Emergency Planning

There have been the inevitable complaints about the lack of a plan to deal with this particular problem. Particularly in the wake of the West Fertilizer tragedy last summer questions have been asked about whether or not Freedom Industries had fulfilled their community notification responsibilities. Ken Ward reported yesterday that the appropriate Tier II forms had been submitted to State and local agencies.

So why wasn’t there a plan in place to deal with this? Why would there be? Emergency planning folks get flooded with Tier II documents, particularly in an area with a large chemical industry like this area in West Virginia. Local emergency planning committees (LEPC) required by EPA regulations are largely voluntary groups with little or no local funding and certainly no federal funding beyond the odd grant or two.

If there is any actual emergency response planning done by these committees (and the Federal government does not actually require any planning to take place otherwise they would be forced to pay for it) it will be concentrated on the most hazardous chemical facilities in the area. In the Charleston, WV area that would include places that make chemicals that will kill people at concentrations of MCHM seen in the current water supply or cause truly catastrophic explosions or fires.

And there is no requirement for anyone to talk to a local water treatment facility in any of this emergency planning. They are not an emergency response organization.

From responses that we have seen to other chemical incidents in the area, I suspect that the chemical response planning and training in the Charleston, WV area is probably some of the best in the country. It didn’t include a response to this type of incident, mainly because it was too low on the potential consequence ladder.

Safe Levels of MCHM Set

Ken Ward reports today that the U.S. Agency for Toxic Substances and Disease Registry (ASTDR), part of the U.S. Centers for Disease Control and Prevention (CDC) had determined a ‘safe level’ of MCHM in drinking water, 1 ppm. This is a calculated value not a tested value. It is from an authoritative source so the local water company will be off the hook for any liability for potential injures from up to that level of contamination in the water supply once things are started back up.

I suspect that they will try to take the levels lower than that, just to be on the safe side. On the other hand, people are getting fed up about not being able to wash themselves or their clothes, and local businesses are being hurt by the necessary closures because of the lack of water. I’m sure the management of the water company is really happy to see this number.