Methane Emissions: Leak Detection Should Be the Last Resort

Photo credit: Google Earth, METEC facilities

Near the foothills of the Colorado mountains lies a replica of an upstream oil and gas production facility. Tanks, well platforms, compressors, and pumping stations, give the impression of a fully operational facility.

It’s an industrial site with a view of stunning mountains in the background.  

However, the facility has no role to play in the production of oil and gas. It is Colorado University’s Methane Emissions Technology Evaluation Center (METEC), which is conducting ground-breaking research on the efficacy of different technologies to help oil and gas companies detect and quantify methane emission leaks from their facilities.

The work is of global importance as scientists and regulators increasingly recognize the potency of methane as a greenhouse gas. Methane is up to 80 times more potent a greenhouse gas than carbon dioxide and its effect is shorter lived.

Daniel Zimmerle is Director of the Methane Emissions Program at Colorado State University and he oversees research at METEC.

In this interview, he offers an overview of the methane detection and measurement technology landscape and explains why he thinks leak detection is one of the worst ways of identifying excess methane emissions from oil and gas operations.

Diana Davis, Oil and Gas IQ: I understand that one of your areas of focus is understanding the various technologies that are available for detecting and measuring methane emissions. Can you tell me more about your work with METEC?

Dan Zimmerle, Colorado University: METEC was originally developed with the idea of doing controlled testing for the ARPA Emissions monitoring program.  We built fake oil and gas well sites and have tested over 50 solutions for emissions monitoring and mitigation. These solutions range from people who have brand new sensor technology being tested in the fields for the first time, all the way through to controlled protocol testing with single blind releases and the like.

In addition to that we do research. For instance, we built pipeline test beds to understand how gas migrates underground. We've done fundamental tests of technologies to understand how gas moves around in the atmosphere close to equipment or close to a subsurface leak. We also spend a lot of time out in the field measuring actual oil and gas facilities so that we can ensure the METEC is representative.

Diana Davis, Oil and Gas IQ:  Could you give us an overview as to the universe of the various technologies that that are being applied for methane mitigation and measurement?

Dan Zimmerle, Colorado University: A good way of thinking about technologies is to think of them as  three pieces. First, there is a sensor of some sort. These days, methane specific sensors are more common but there are still some VOC (Volatile Organic Compounds) sensors around that measure a range of gases.

The second piece is the deployment aspect. You have to get the sensor in a position where it can monitor the site. There is a plethora of options for this. Sensors can be mounted on a post or put on drone, car, aircraft or even on satellites. The salient point here is that there is a deployment process that brings the sensor in contact with the site.

The third piece is the analytics. The sensors take measurements and the analytics must translate those measurements – effectively a time series of values - into something useful to allow for detection of emissions.

You can fill in your menu with every possible sensor type combined with every possible deployment type. There are cameras on drones, and cameras that people hold, and cameras that you put on vehicles. There are line lasers that are stationary, line lasers that move, and droned mounted line lasers that point at the ground. Every combination is possible.

The sensor technologies are still evolving but the deployment methods are generally mature and generally use a certain number of recipes.

The analytics piece, however, is still a big challenge. There is a lot of research going on to address the issues but we’re not there yet.

Diana Davis, Oil & Gas IQ: What are what are the issues with analytics that oil and gas companies are confronting?

Dan Zimmerle, Colorado University: It’s worth noting that there are slightly different challenges from the oil and gas perspective versus the challenge that solution developers are confronting for this.

Let's back up first and talk about how the old style of leak detection differs from today. In the old style of leak detection you held a sensor in your hand, you got very close to something and you measured it. You could also look at something through a camera.

The person who was checking for leaks were up close and personal with the equipment; they would record their measurements and run diagnostics.

However, the new generation of solutions is looking for the leak after it's being transported by the wind. Nobody is up close and personal. Mostly, these solutions are not being operated by a person most of the time.

Instead, they’re taking an image or a measure of a gas concentration and trying to draw some conclusions. That's really what the analytics are doing. They are extracting information about the leak from gas that's been transported in the wind.

We know a lot about how gas gets transported in the wind at the distances of hundreds of meters to kilometers to hundreds of kilometers. That's a fairly well-developed science.

On the other side of the spectrum, you can get very compact computational fluid dynamics (CFD) to understand how the gas mixture is going to move in a very small space.

The challenge for the industry is that methane emissions occupies a tricky middle space. We're a little too close for dispersion models to just plug in and work, and we're a little bit too complex and it's a little bit too expensive for a CFD model to work. In any case, CFD models only give you an average answer anyway. They don't give you a look at what is happening now and that’s what is really driving this whole problem.

The analytics for methane measurement and detection occupies this messy near field space that is difficult to analyze.

Diana Davis, Oil and Gas IQ: Are there any particular emerging technologies or approaches that you think might be promising for methane detection and reduction?

Dan Zimmerle, Colorado University: As a test center we don’t want to back specific solutions because we want to maintain our neutrality. However, there are some general statements about the bright spots for different types of solutions.

There are two big classes of technology. First, you have surveying solutions. These typically are human operated. You go to the site or fly over it. But it’s a survey of a particular site at a particular point in time. It only happens once in a while.

There are a number of different surveying approaches. There is the classic optical gas imaging. There are interesting things going on with drive around sampling and with aerial overflights. Even satellites are being used.

Each of these methods have different detection thresholds, and different weaknesses and strengths. The general idea is that survey solutions are expensive but you do them with high quality, high precision equipment and, therefore, you expect more accurate results. The challenge is that you only get those results periodically.

On the other side, you've got continuous monitors. You set something up and it runs as much of the day as it possibly can. That's where people want to go longer term. They want these continuous tonitors to work on their sites because that's probably the best way to knock down large emissions so that they are caught as soon as they start.

These solutions are more autonomous, which means there's more software. There's more analytics. They're cost sensitive because they have to be out on that site and running all the time so there is still more development work to do on these at the moment.

Diana Davis, Oil and Gas IQ: Is that the reason that automated methane emissions monitoring aren't more widely adopted?

Dan Zimmerle, Colorado University: There are definitely useful solutions out there now in terms of performance per quantification and detection. However, they’re perhaps not quite as tight on the quantification as people would like. The detection limits for finding leaks can be quite high in some cases.

You can think about these technologies as being not quite teenagers yet. They’re in middle school and they still have some ways to go before maturity.

But, that being said, I think they’re doing well for the age of the products.

Diana Davis, Oil and Gas IQ: At METEC you’ve tested over 40 leak detection and quantification technologies. With so many options available how can oil and gas operators go about effectively evaluating what mix of technologies might work for them?

Dan Zimmerle, Colorado University: Sometimes our aspirations for what the technology should do outstrip the capabilities and that leads to all kinds of disappointment. First, there needs to be a meeting between what's possible and what you're trying to do.

The second point is that it’s important for operators to try the solution and do some robust tire kicking. Don’t assume the data sheet is right. Test it to see how it performs in your system.

For instance, let’s say that I'm an upstream operator and I'm concerned about emissions from my tanks. There are a couple of different things I could. One of them is to put in point sensors that would look for emissions down wind from those tanks and signal if there was something large going on. That would give me an indication that I should dispatch people when the detector hit a certain level.

Another option, I would have is to put a camera on the site. This could be a permanently fixed gas camera – there are lots of different ones available – that would signal to me when it saw emissions coming out of a tank.

My third option would be to put a tank pressure monitor on. If that tank pressure doesn't stay in the right realm then you can get somebody out to the site to make sure that it's not leaking.

Each of those solutions is a different flavor of what it takes to solve the same problem, each with advantages and disadvantages.  

For instance, the point sensor might find problems that are away from the tank but on the same site. However, there could be lots of false alarms. Or, when people arrive on site it can be hard to trace exactly where the emissions are coming from.

The camera might tell you exactly where the problem is - you can literally see it in the picture – but the camera might not work at night and it is blind to what’s going on behind it.

With the tank pressure monitor, you'll never know whether there was an emission or not, but you'll have immediate feedback that something has caused the tank pressure to change.

So, people should think about what they're trying to accomplish and then look at not just the leak detection system but a more comprehensive protection system.

I often say that leak detection is the worst of about five ways of reducing methane emissions. It's roughly the equivalent to detecting overloads on your household circuits with smoke alarms. It comes a little late in the process.

In our household we put a sensor on the circuit. There are other things oil and gas operators should be doing with regards process monitoring and site design. Leak detection is the smoke alarm.

Diana Davis, Oil and Gas IQ: What are some of the things that that operators could be thinking about implementing to catch methane emissions a little bit further upstream and the process?

Dan Zimmerle, Colorado University: There are some obvious ones. I’ve already mentioned monitoring pressure in the tanks. You can also add a thermocouple or other type of sensor on both the pilot light and the combustion chamber of flares.  

You can also track emissions in your equipment such as rod packing emissions or crankcase emissions. You can have monitors to find if there's an unusual amount of emissions.

Some companies are looking now at where their problem children are and going after process monitoring or changes to address them. This might involve swapping out equipment such as gas-powered pneumatics, training people or putting in place process monitoring.

Diana Davis, Oil and Gas IQ: What advice would you give to oil and gas operators who are looking to reduce their methane emissions?

Dan Zimmerle, Colorado University: If you think about the life cycle of a site, start with designing your new sites with EHS considerations. Don’t try retrofitting them on afterward as that has proven itself to be exceptionally difficult and expensive.

Then, in addition to site design, we need to be more critical about the equipment we purchase, and start to do targeted replacement of existing equipment. Process monitoring can alert you to when the site is behaving differently from how you have designed it, allowing you to make appropriate interventions. Finally, you can do some type of leak detection.

If you're trying to mitigate emissions and you have a bunch of existing sites, I would encourage people to to look at leak detection as a way to find and fix problems, but to also consider process monitoring. On large sites, it may be more effective to do more process monitoring and on smaller sites, leak detection might make more sense.

The path you choose has to fit with the operating style of your company.

Diana Davis, Oil and Gas IQ:  What are you looking forward to at our upcoming Methane Mitigation Summit?

Dan Zimmerle, Colorado University:  I go to these events to meet people. It’s great to get people together; it’s quite energizing for everybody in the industry.

Interested in Learning More About this Topic?

Dan Zimmerle will be speaking at our upcoming Methane Mitigation Summit taking place in Houston this December. Get up to the minute updates on methane mitigation research and development efforts including monitoring sensors, data management systems, modular conversion technologies and alternative uses for stranded natural gas. Download the agenda for more information.