<<BACK TO HOME

Stricter control of NOx emissions will require more sophisticated aftertreatment systems and better simulation of tailpipe gas. By Tristan Honeywill

NOx treatment is going to become a bigger issue in the industry. National emission limits will also start to gain more prominence from 2010 and this is likely to focus attention on NOx and the car industry. It’s unlikely there will be much understanding or sympathy toward how problematic the trade-off in emissions is.

Any measure to reduce CO2 emissions will inevitably put more pressure on the NOx aftertreatment: cutting CO2 automatically increases NOx emissions from the engine. Diesels may produce less CO2, but current particulate technologies use oxidation catalysts that remove carbon monoxide and hydrocarbons and, in doing so, cause more NO2.

To make this technology cheaper, some OEMs use a small oxidation catalyst in front of a coated particulate filter. NO2 is useful in burning off the soot in the filter, but can pass through to the tailpipe. Any NO passing through the oxidation catalyst can be also converted into NO2 by the coated filter and emitted.

It’s a vicious circle. The higher the NOx emissions, the better the fuel economy is. Adding bigger oxidation catalysts to reduce the carbon monoxide then increases the NO2.

Catalysts are relatively simple to do: they don’t require urea injection and its associated extra components. The disadvantage is that fuel is used to optimise the tailpipe-out emissions. Anything that increases fuel consumption is unlikely to be popular in future.

Urea injection and NOx adsorption will become more important if diesel engines are the mainstay of CO2 reduction strategies. Deciding which approach to take will be linked to the vehicle weight and engine size, but in general it will be NOx adsorbers for the smaller engines, selective catalytic reduction (SCR) for the bigger units.

OEMs will also have to look at their strategies for lean-burn gasoline engines. If these require a lot of costly aftertreatment that saps the powertrain’s efficiency, there may be question marks over their viability. Retaining a lambda-one engine that can run with an efficient three-way catalyst might be more sensible in a lot of cases.

There’s usually a fuel economy penalty for any kind of aftertreatment. A NOx adsorber must be regenerated every few seconds and that means switching to rich mode in order to decompose the catalyst’s barium nitrates. During this rich phase it is possible to create ammonia on precious metal catalysts.Hard acceleration also produces rich conditions – purging is then automatic.

OEMs will have to balance the engine and aftertreatment together, but extracting more performance from aftertreatment systems without increasing packaging size or cost will be a top priority. For urea injection systems, improving the dosing and evaporation will be important. The smaller the droplets, the larger the surface:volume ratio is, and the better their evaporation.

“When you inject urea, you have droplets that aren’t small enough to evaporate immediately,” says Dr Jörg Oesterle, a development manager at exhaust system supplier Eberspächer. “Their mass and inertia means they can’t follow the flow around bends, so you achieve only a fraction of the volume in an SCR catalyst.”

It is important to prevent the urea from ending up on the walls of the exhaust. They can evaporate here, but mixing them in with the main flow of the exhaust gas is difficult. It can still be made to work, but it is a relatively inefficient way to feed the SCR catalyst.

Eberspächer is looking at ways of lengthening the exhaust so there is more time for the urea to evaporate before it enters the SCR catalyst. But it thinks secondary devices that can atomise the droplets will be the best solution.

“We’re thinking of mixer plates,” says Oesterle. “When the larger droplets hit the plate, you get smaller droplets. And they also provide a surface for evaporation.”

Packaging the systems and getting more cells per square inch into the catalyst is the main issue, says Emitec. “Increasing cell density and reduced wall thicknesses have contributed a lot already,” says chief operating officer Wolfgang Maus. “But laminar and turbulent flows have theoretical differences in efficiency of a factor of between five and 20. The question is how much can be achieved realistically.”

Maus says that with current coating technologies, it is possible to reduce the SCR catalyst’s size by some 30 to 50 per cent. But the real advances come from turbulent flows. Laminar systems work by transferring the emissions to the wall by diffusion. Because this relies on concentration differences, it is slow.

“Turbulence is faster,” says Maus. “It forces the reaction. Pollutants come to the wall faster and the coating has to react faster. It has to have different porosity and grain structures.”
To be more reactive, the coating has to be thinner. Its contribution to space-saving is significant. Emitec will put the first of such systems into production later this year. If greater NOx reduction were required the future, this could be used instead to increase efficiency.

Urea is not normally injected at full load, so mass flow rates are not high enough to create turbulence. It could be possible to introduce devices for this effect, but it would depend heavily on the strategy, says Eberspächer: the injection point, operating points, temperatures and mass flow rates.

“In some ways it makes no difference if you are talking about SCR or NOx adsorber applications,” says Rainer Lehnen, Eberspächer’s commercial vehicles director. “The aim is to bring ammonia or hydrocarbons and NOx to the catalyst at the same time to get the highest performance. It’s the bended metal that creates a uniform flow.”

In the commercial vehicle sector, the question gets even more complex. For one application it could be a mixer, for another just a complex flow created by different chambers and pipes.

And as onboard diagnostics become more important to the engine’s operation, the developers have to make sure temperature sensors can see a good mix of the complete flow, not just that of a single cylinder firing.

Simulation work is helping, but more sophisticated tools will be needed. Eberspächer uses standard CFD codes to ensure homogeneous distribution of the flow over the cross section of the monolith.

It works well for single-phase flows such as exhaust gas or hot air. But SCR is a two-phase flow – there is the injected urea and the hot exhaust gas – and the normal commercial codes are not good enough to cover all the features involved. Flow impingement on to the wall, atomisation, evaporation and the chemical reaction on the monolith are all too complex.

“At the moment these are separate calculations, but we’re working on a validated simulation tool that can do this for customers,” says Oesterle. “It’ll be done in steps: first we’ll deal with atomisation and evaporation, then the wall flow, then the chemical reaction.”

Emitec’s immediate concern is the diffusion and reaction of the pollutants with the washcoat. “Simulating the flow is not enough,” says Maus. “We’ve already applied common codes to our catalysts’ flow dynamics and thermal dynamics. The issue is understanding distribution and grain sizes. We’re working on this with universities and trying to find out what the correct approach here is.”

It will be several years before OEMs have usable tools at their disposal. It would be useful if their introduction could coincide with renewed legislative efforts to reduce NOx.

If they are to be available sooner, then OEMs, suppliers and universities need to start working closer together on this now.