Jessica Rowbury looks at the various light-based techniques for monitoring atmospheric pollution
Within the last two months, several remote sensing technologies have been demonstrated that may help in the struggle to reduce atmospheric pollution. In June the UK’s National Physical Laboratory launched an upgrade of its mobile lab, which uses differential absorption lidar (DIAL) to measure gas emissions from power plants and search for leaks at industrial landfills.
And, researchers from Örebro University in Sweden have created a mobile robot, capable of detecting methane leaks in streets, mines and landfills. The Swedish team is considering using the robot as part of a US government scheme launched in April to investigate why methane levels in the USA have been underestimated by up to 75 per cent.
However, it is not just about detecting and measuring pollutant gasses, but ensuring that companies are reporting accurate levels. As described in a paper published in May, the Los Alamos National Laboratory (LANL) in the USA used a remote sensing system to measure gas emissions from the two largest power plants in America over three years, and was able to verify that they matched the official figures reported to the Environmental Protection Agency. The technology has potential in the future to be used as a global emission monitoring system to verify international regulations on fossil energy emissions.
The DIAL lab developed at the UK National Physical Laboratory (NPL) can measure emissions from up to 3km away, and can produce 3D emission maps and the mass emission rates for airborne pollutants such as hydrocarbons. The mobile lab has already been used for monitoring emissions in the USA − including refineries, landfill and coke works − and in a number of other European countries. The updated system launched in June has improved capabilities, including better sensitivities and improved analysis software so that 3D plots can be produced while in the field, as opposed to after the measurement campaign.
The system works by sending out two laser pulses and analysing the scattered light that returns. The two light pulses are fired into the atmosphere sequentially; one is tuned to the wavelength absorbed by the specific gas being measured, and the other pulse acts as a reference. ‘We use two wavelengths that are very similar, so that only the difference between the signals is that one will be absorbed by the gas we have tuned to,’ explained Rod Robinson, principal research scientist at NPL and a team leader of the DIAL project. The difference between the two returning signals is then compared to determine the gas concentration.
A concentration map is created, based on multiple measurements. ‘With time, we are getting the scatter from further away − so we are getting a range-resolved signal,’ explained Robinson. ‘We then look at the difference between those two signals; we get concentration but then we also get the distance information. And we will basically fire out a whole sequence of these pulses to build up a map of concentration in the atmosphere.’
By combining the concentration map with wind speed data, a mass emission rate for an industrial site can be determined. ‘Because you get a map of the concentration, we can do a vertical scan and get a cross section of a plume. We can combine that with the wind speed data which gives you the flux − it gives you an emission rate,’ said Robinson.
These features also provide useful tools when searching for gas leaks. ‘For landfills, we can scan horizontally, so you can build up a horizontal map that will show hotspots or areas where there isn’t good containment or [the landfill company] aren’t capturing the methane,’ Robinson added. ‘So, we do work for landfill companies where they can use that information to plan repairs on the landfill.’
Identifying emission leaks not only reduces a company’s environmental impact, but can also deliver clear commercial benefits. ‘We certainly have customers at the moment who use our services – particularly with methane, which is a valuable commodity in itself,’ Robinson said. ‘For example, in a landfill site, companies will be trying to capture the methane to not only reduce their greenhouse gas emissions, but also because they can capture it and put it into a gas engine and generate electricity with it − they are basically turning that potential waste gas into an asset.’
The environmental impact does have to be noted, however. ‘Methane is a very powerful greenhouse gas − it is 56 times more powerful than CO2,’ said Achim Lilienthal, leader of the Gasbot project at Örebro University. Following a study published earlier this year in Science demonstrating that methane levels in the United States have been significantly underestimated, the US government has called for ways to detect potential unknown gas sources and leaks.
‘Since February [when the study was published], it has been known that the methane concentration over the US is much higher than the Environmental Protection Agency knows,’ said Lilienthal. ‘It is not yet clear, but the measured concentration seems to be higher by 50 to 75 per cent. Therefore, there was a special report by the Intergovernmental Panel on Climate Change (IPCC), a group associated with the UN, and responding to that, in April, the Whitehouse launched a special programme to find out where the gas comes from.’
The team of researchers at Sweden’s Örebro University are planning to participate in US projects relating to this government scheme with its remote gas sensing device, the Gasbot. The robot, which was presented at the 2014 IEEE International Conference on Robotics and Automation in Hong Kong in June, has been designed to travel through landfill sites and mines to detect and map methane leaks from up to 30 metres away.
The system uses the general principle of tunable diode laser absorption spectroscopy (TDLAS), and works in a similar way to the DIAL lab in that two laser beams are sent out, with one of the rays tuned to the wavelength of methane, and the other beam acting as a reference. The light that is reflected by the environment and detected by the sensors is then measured. The sensors in the system are commercially available, and are also specific to methane.
Through tests carried out at a landfill site, the Gasbot was able to detect methane in concentrations of down to five parts per million in the air, and it is also capable of producing a concentration map. In the same way as the DIAL lab, the gas distribution is not determined immediately, but worked out after many measurements are taken and analysed. ‘What you know from a single measurement is that there was a certain total absorption, but the single measurement doesn’t give you the length or how far away the beam was reflected, but it gives you the total absorption,’ explained Lilienthal. ‘This is something you need to find out by taking many measurements which is what we are doing.’
At the same time as the measurements are being made, the software builds up a 3D model of the world, which allows the team to understand how far the laser beam was reflected. ‘This we do for every measurement − for every measurement we then know, up to some [degree of] error, exactly how the laser ray proceeded in space, and we know the measurement value,’
Lilienthal said. Then, through the use of algorithms, the most likely distribution of the methane in the air is determined. ‘We discretise the environment into small cells and we make some assumptions about these cells,’ Lilienthal added. ‘And then we have a method in order to find out: “What is the most likely explanation for our set of measurements? What is the most likely distribution of gas to explain the set of measurements that we have?” This is the core of the algorithmic approach.’
Spectroscopy for gas measurements
Other instruments using spectroscopy have been developed to measure pollutant gasses remotely. A spectroscopic system developed by the Dutch National Institute for Public Health and the Environment (RIVM) has managed to overcome traditional challenges of measuring atmospheric ammonia. ‘The instruments available on the market all have the problem that they use sampling lines − so they sample the air through a tube to bring it to the instrument,’ explained Hester Volten, air quality scientist at the RIVM. ‘Ammonia is very sticky, so it gets stuck to tubes, filters, or any surface, which causes problems.’
The MiniDOAS system is based on differential optical absorption spectroscopy (DOAS) and consists of a UV lamp with a wavelength range of 200-230nm, and an Avantes spectrometer. It has the advantage that no ammonia comes into contact with any part of the machine because the measurements are taken from an open path of 15 to 20 metres. The light is reflected into the spectrometer by a retroreflector, and the amount of light absorbed at certain wavelengths is proportional to the amount of ammonia gas present.
The team at RIVM are now in the process of introducing the MiniDOAS into the Dutch National Air Quality Monitoring Network, as ammonia forms a major environmental problem in the Netherlands.
Volten commented: ‘We started to make the system smaller and less expensive, and this is where we brought in the Avantes spectrometer, as the spectrometer was the limiting part in the instrument that made it really expensive. By using a smaller Avantes spectrometer we could downscale the whole instrument and produce it a lot cheaper and make it available for ammonia monitoring.’
The MiniDOAS replace a system used by the RIVM that currently needs €15,000 worth of maintenance per year, including the use of various chemicals. Volten believes that the miniaturisation of spectrometers and sensors has allowed for cheaper and easier pollutant gas monitoring, and this trend will continue in the future: ‘It is really exciting to see how sensors and other instruments such as spectrographs are getting smaller,’ she noted. ‘I think that will open up lots of possibilities − to have sensors that are smaller and at lower costs so you can widely employ it − to get a better view of the spatial distribution of pollutants.’
Dr David Creasey, vice president of sales and marketing at Ocean Optics, agrees that the miniaturisation and reduction in cost of spectrometers has opened up more ways in which the environment can be monitored, and this trend will continue: ‘A classic case of it is that Halma − our parent company − has just formed an environmental and analysis group, of which Ocean Optics is a part. That is a very clear indication of how serious environmental monitoring using spectroscopy is. We are investing in that area, because we believe that area will grow.’
Spectroscopy is being used to measure pollutant gasses by Ocean Optics’ customers in a variety of ways, according to Creasey: ‘Different customers are using spectroscopy in different ways.
There is a global network of environmental scientists looking at trace levels of pollutants in remote locations across the globe, in inner cities, on land, on sea and in the air,’ he said. ‘There are also a growing number of companies using Ocean Optics modular spectroscopic tools to provide real-time pollution monitoring. Typically NOx and SOx concentrations are measured from power station stack emissions and dense urban environments in order that the ever increasing legislation of these pollutants can be policed.’
In terms of where he sees this environmental sector moving towards in the future, Creasey noted that the volume applications will be driven by this legislation: ‘For example, new maritime pollution regulations to be introduced in the next few years will demand that emitted levels of NOx, SOx and a good many hydrocarbons are to be significantly reduced, meaning that this industry will drive a boom in environmental monitoring for the remainder of this decade.’
But, currently, there are few ways of verifying that international legislation regarding gas emissions is being adhered to. This was the aim of a study published in May by a team of scientists from the Los Alamos National Laboratory in New Mexico. The researchers spent three years measuring greenhouse gas emissions from two coal-fired power plants in the Four Corners area of northwest New Mexico, the largest point source of pollution in America, to verify whether they matched the figures held by the Environmental Protection Agency (EPA) to comply with the Clean Air Act.
The researchers also wanted to ensure that the in situ measurements being taken by the power plants were accurate and not affected by atmospheric factors. ‘What we wanted to do was try to ask the question of if we can look at larger scales − kilometres and tens of kilometres scales − and get the average air quality, rather than these point measurements which are affected by metrological mixing and transport,’ explained Mavendra Dubey, who led the project at LANL. ‘We wanted to get a larger scale measure of the pollution and get it remotely.’
Two spectrometers were installed at a site approximately 10 kilometres away from the power plants, and they used the sun as the light source. ‘They use the sunlight to make the measurement − so they look directly at the sun and track the sun from sunrise to sunset − the sun is the source of light, and you collect very high-resolution spectra,’ said Dubey. ‘Every two minutes you get the spectra, you analyse the spectra for all of the gases because they have unique absorption signals, and then you measure column concentrations of the greenhouse gasses.’
After three years of measurements, the team were able to verify the results that the EPA held. ‘In-stack monitoring of power-plant emissions is mandatory in the United States, and they are reported to the EPA to comply with the US Clean Air Act. So we ran a model with those initial specifications and conditions, because we knew the answer, and then we did a forward model and the model was able to reproduce our measurements. This meant that our verification technology was quantifiably accurate,’ explained Dubey.
In the future, this could allow for a global emission monitoring system, for the verification of inventories and reduction of emissions claimed by individual nations, Dubey noted: ‘One of our big project motivations was to answer the question whether we could verify pollutant levels. For China there are huge gaps, where even their national and provisional inventories don’t add up. CO2 is a very important issue and we want to make sure that everybody is reporting accurately and that there is a good inventory,’ he concluded.