Integrating Ultra-Low NOx Technology Presents Challenges
At the recent SAE COMVEC Digital Summit, experts at Navistar, Cummins and AVL discussed potential solutions to meet upcoming ultra-low NOx regulations.
Ultra-low NOx regulations proposed by the California Air Resourced Board (CARB) call for 0.05 g/hp-hr tailpipe NOx in 2024, with a second stage in 2027 further reducing NOx to 0.02 g/hp-hr. A new low-load cycle will require 0.2 g/hp-hr. Such stringent standards will lead to an increase in cost and complexity for OEMs, explained Navtej Singh, chief engineer for powertrain systems at Navistar during a technical session at the SAE COMVEC Digital Summit. “The solution we’re trying to create, we would like to see how it can live for another 5 to 10 years because the investment – we’re talking hundreds of millions of dollars,” he said.
To satisfactorily meet the stricter standards, significant improvement in the durability and robustness of sensors, catalyst and dosing technologies is still required, Singh emphasized, and more accurate sensing and precise actuator systems also are needed for 2027+ emission systems. “Navistar is still looking for more accurate solutions in NOx sensor technology – for example, a NOx sensor other than ±10 percent. We’re expecting to see ±3 percent,” he said, adding that relaxation in OBD (onboard diagnostics) is also key to early adoptions.
Singh was joined on the COMVEC panel by Lisa Farrell, director of Cummins’ Accelerated Technology Center, and Gustav Johnson, chief engineer at AVL Powertrain Engineering. The three experts covered a range of topics during the virtual session.
What’s the biggest challenge to meeting ultra-low NOx?
Singh: Thermal management is one of the biggest challenges for us, specifically when we consider NOx control in the cold FTP [Federal Test Procedure] and low-load cycles. We looked into different avenues, how we can use electrically heated catalysts or substrates, how we can enhance the aftertreatment heating by doing the exhaust bypass, we looked into cylinder deactivation [as part of the SuperTruck program], and we found pros and cons of each application.
But one of the biggest challenges we’re still facing is how to control the engine-out NOx that is being produced in the first 60 seconds – we are still working on these challenges. We can control hot FTPs pretty well; we are still struggling to find a reliable system that can be robust to 1.2 million miles and can meet the low-NOx specifically in the cold cycles and the low-load cycles while meeting GHG standards.
What technologies make sense for low-load operation?
Johnson: When we talk about the ultra-low NOx regs, there’s the low-load cycle and there’s the cold FTP and the hot FTP, and you need different approaches for all of them. As Navtej mentioned, the hot FTP isn’t much of a challenge assuming you have everything there for the others. The cold FTP is all about getting hot quickly, and the low-load cycle is about staying hot. The technologies that allow you to get hot quickly aren’t necessarily the same technologies that allow you to stay hot efficiently.
Cylinder deactivation [CDA] is a good example of that – it makes a lot of sense for staying warm, it’s not necessarily helpful for getting warm, as an example. There’s also a CO2 angle of it, as well. You need to get that heat for as little extra CO2 penalty as possible, so that’s going to be the key metric.
The last thing I’d add is model-based controls – you can’t wait for there to be a problem to respond, you need to see it coming in terms of what the catalyst is going to do 10-15 seconds into the future from a temperature point of view. All of that will need to be integrated to keep that temperature in that active range for the SCR [selective catalytic reduction] catalysts.
What set of technologies are required to meet the 2027 NOx standards?
Farrell: As we work toward 2024 and meeting 0.05, we’re going to learn what it’s really going to take. The challenges are not just FTP and RMC-SET [Ramped Mode Cycle-Supplemental Emissions Test], there’s the new low-load cycle and that lab test is probably not even the greater challenge. The in-use requirement and the temperature requirement down to -7°C and being compliant to that are going to be extremely challenging for these technologies [like CDA, close-coupled SCR, and the e-heater].
As we are developing the products, we’re going to find out what it takes to get to 0.05 and then what do we need for 0.02. We’re going to have to maybe error on the side of protecting for some additional options if we need them. There are just some things that we don’t have enough experience even at 0.05 yet to say the in-use requirement is going to be met by any of those technologies for 0.02.
Is existing DEF dosing technology appropriate for close-coupled SCR catalysts?
Singh: Absolutely. There are two kinds of prevalent dosing technologies available, one with Bosch and one with Cummins. I think both technologies are feasible. We believe, based on the data we have, that controlling the dosing and trying to dose at a lower temperature doesn’t make much difference. What we believe is that the challenge is how to heat up the system that is ahead of it. That is a big challenge, because even if you start dosing early, if your catalyst is not hot, reactions will not take place. But I don’t want to discourage the manufacturers who are making the new dosing technologies – we would encourage them to keep developing.
What are the control implications for meeting low-NOx standards?
Johnson: The main thing to keep in mind with controls for a lot of these [solutions], there are some low-level things like any increase in sensor accuracy is going to make diagnostics and similarly engine-out NOx decisions less of a competition. So that’s on the low level – precision of the individual components. That is an enabler, it’s by no means a solution to the challenge.
I think the real ‘magic,’ so to speak, is going to be model predictive control. So not just reacting to what’s happened and adjusting, and not even looking ahead a few seconds into the future, but really trying to understand potentially integrating with some advanced features on the vehicle to understand what’s coming [further out] and getting ready for it. So that you’re not trying to dig yourself out of hole, but step around the hole before it comes.
Does Cummins see a hydrogen ICE in its profile?
Farrell: It’s well known that Cummins has invested significantly in hydrogen technologies, particularly around fuel-cell technologies with the acquisition of Hydrogenics. For 2024, a hydrogen IC [internal combustion] engine in California is not a prime path for Cummins. I certainly think that for European regulations it’s being evaluated much more heavily. But all hydrogen technologies will continue to be evaluated as potential options for future regulations and products for Cummins.
Is electrification the ultimate solution?
Singh: Only time can predict the demand of electrification in the trucking industry. For the heavy-duty Class-8 application, adoption of e-mobility has challenges ahead of it. Government involvement and incentives will be key in adoption. In North America when we sell the trucks to customers, to a fleet of like 10,000 trucks, think about a scenario that they need a 1,000-kWh battery system. One battery is running at about 2.2 kWh per mile, that will give them about 500 miles. So basically we’re talking 10 million kWh of energy needed by one company per charge – think about what kind of a charging infrastructure the U.S. needs to provide them power.
We do see other applications where electrification is going to take over faster like bus and vocational applications – school buses are ideal because you drop off the kids, you pick up the kids, you have a 5-hour window you can charge them up. But for freight, if you’re asking specifically on the heavy-duty application, I think IC engines are going to stay longer than expected.
Farrell: We’re in the process of putting out early learners now and hopefully within the next few years we’ll have a better understanding. But I don’t think we know that right now.