Conference report: Sensors for Exhaust Gas Cleaning and CO2 Reduction
20 July 2015
The 2nd International Specialist Conference: SENSORS for Exhaust Gas Cleaning and CO2 Reduction was held in Nuremberg, Germany on June 23-25, 2015. The event—organized by SV Veranstaltungen—started with an introductory seminar on automotive sensor technology. The technical program in the following two days included 16 presentations by experts from the industry, academia and NGOs on exhaust emission testing, various types of sensor technologies, and integration with onboard diagnostics (OBD) systems. The conference program also included a tour to HE System Electronic, an electronics and sensor manufacturer, where delegates could see production technologies used in the semiconductor industry, that are now utilized in many types of modern automotive sensors.
The introductory seminar, given by Stefan Carstens [EngineSens Motorsensor], started with an overview of sensors used for engine and aftertreatment control, illustrated on the example of the new Audi V6 TDI engine for the European and the US markets. The following parts of the seminar covered the principles, design and operational issues of several types of sensors, including exhaust gas temperature resistive probes (Pt200); NTC (negative temperature coefficient) thermistor based temperature sensors; thermocouple probes; oxygen/lambda sensors; NOx sensors; pressure sensors; and soot sensors.
Emission Testing. The technical sessions opened with a talk by Prof. Reinhard Kolke [ADAC] who presented a wider perspective on in-use emissions from passenger cars in Europe. Germany’s ADAC—the second biggest motor club in the world—has been among the first organizations to conduct in-use emission measurements and to report on high levels of NOx emissions outside of the NEDC test cycle from many diesel car models—an issue that has now been widely publicized. One of recent ADAC studies compared emissions from three Euro 5 cars: Mazda 6 2.2 Skyactiv-D with in-cylinder NOx control; BMW 320d BluePerformance with a NOx adsorber system; and VW CC 2.0 BlueTDI with urea-SCR aftertreatment. The vehicles were driven for 18 months over a distance of 100,000 km, and periodically re-tested over the NEDC and the ADAC highway cycle. While all three cars met the NEDC limits over the 100,000 km, over the highway test only the VW vehicle with SCR produced low NOx emissions, comparable to the NEDC levels—the other two vehicles produced highway NOx levels three to five times higher. NOx emissions from the BMW with the NOx adsorber showed a noticeable deterioration over the distance of the test, while the Mazda NOx emissions slightly decreased over the duration of the test.
Real driving emissions (RDE) from the three vehicles were also tested in the field using PEMS equipment. The RDE results showed that the VW SCR system was generally effective at speeds above 40 km/hr, but the efficiency dropped to very low levels below 30 km/h. The VW vehicle consumed about 1 L of urea solution (AdBlue) per 1,000 km, which was higher than the manufacturer specifications. As the AdBlue tank volume was only 17 L, AdBlue refills were necessary between the regular service intervals. The refills were inconvenient and cumbersome because of the location of the AdBlue tank, with access from the vehicle trunk. It was suggested that the AdBlue refill nozzles should be placed outside of the vehicle and AdBlue pumps should be introduced for passenger cars, similar to those used for heavy vehicles.
To better control in-use emissions, particularly diesel NOx emissions, the EU is adopting new regulatory emission tests. The NEDC test cycle will be replaced by the worldwide harmonized light vehicles test procedure (WLTP)—designed to better represent European driving patterns—and RDE testing in real traffic conditions using PEMS analyzers will be required in addition to laboratory testing. TÜV SÜD provided an update on the implementation of the new tests [P. Mast, presented by S. Carstens]. While the schedule for the changes has not yet been finalized, it is expected that the switch to WLTP could occur over the period of 2017-2019. Initially, WLTP results would be used for vehicle labels, while NEDC would continue to be used for regulatory compliance testing. This transitional, dual cycle period will create a significant increase in demand for testing capacity by vehicle OEMs (the WLTP test is also longer than the NEDC), creating business opportunities for providers of testing services.
The WLTP test is not expected to create technical challenges: the test is hotter, with a less pronounced cold start period, which is beneficial for catalyst light-off. The faster warm-up and longer duration may also produce a CO2 emission reduction. However, since the WLTP includes less idling, engine stop-start technologies will be less effective compared to NEDC testing.
More technical challenges are expected from the RDE testing, where emission standards will have to be met under a wide range of operating conditions. The expected changes in emission technology include larger catalyst volumes, replacement of NOx adsorber systems with SCR, and the use of higher EGR rates over the entire speed range. In gasoline engines with three-way catalysts, mixture enrichment strategies will have to be limited or entirely eliminated, and catalyst protection strategies will likely rely on exhaust temperature control via cooling, using engine block-integrated cooled exhaust manifolds.
Oxygen and NOx Sensors. Development trends in the are of oxygen sensors were discussed by Bosch [Petra Neff]. Commercial oxygen sensors, as well as the related NOx sensors, utilize conventional ceramic sensor technology. In its simplest form—the “switching” or two-step O2 sensor capable of distinguishing between a rich and lean condition (Bosch LSF)—the sensor consists of a ZrO2-Y2O3 substrate with two Pt electrodes, one exposed to reference air, the other to the measured exhaust gas. The oxygen concentration is determined based on the Nernst voltage between the electrodes. A more complex design is used in wide range sensors (Bosch LSU) that can measure the actual O2 concentration. The sensor includes an internal cell in the zirconia substrate, where voltage is applied to pump out oxygen ions. The pumping current, compared to a reference current, is a measure of oxygen concentration. The sensor incorporates a heater to bring it quickly into the operating temperature range. While this technology can provide good stability and robustness, the sensors are still vulnerable to thermal shock from liquid droplets impinging on the hot ceramic surface. Other development targets include low power demand and faster readiness after engine start-up. The latter demand is driven by the upcoming emission regulations with a focus on cold start emissions. Early start of closed-loop λ control in SI engines can bring significant benefits—in a SULEV application, advancing the start of closed-loop λ control from 16 s to 8 s can reduce NMOG emissions by about 20%. To address these challenges, Bosch developed a sensor with a power-optimized heater, capable of quick heat up of the electronics within 5 s. A porous thermal shock protection (TSP) layer was introduced to provide better protection from liquid droplets impingement.
Commercial NOx sensors also utilize an Y-stabilized ZrO2 (YSZ) electrochemical electrode and share a similar operating principle. In the NOx sensor, NO2 is reduced to NO, and oxygen ion pumps are used to remove all oxygen from the system, allowing for the measurement of the nitric oxide. Hence, compared to the oxygen sensor, NOx sensors are more complex (the UniNOx sensor, for example, includes three oxygen pump cells) and the voltage signal is orders of magnitude lower. The remaining issues—including thermal shock vulnerability, poisoning by magnesium [S. Carstens, EngineSens], cross-sensitivity to ammonia, the lack of NO/NO2 selectivity, accuracy (especially at low NOx concentrations), as well as slow response—drive the ongoing development of alternative NOx sensors.
UST Umweltsensortechnik [O. Kiesewetter] presented their UST Triplesensor technology—a ceramic chip with Pt microstructures and three MOX layers, applied using semiconductor production technologies. The three MOX layers are sensitive to compounds of varying redox character. The capabilities of the sensor were illustrated with data on simultaneous measurement of ammonia, SO2 and VOC. Preliminary laboratory results were also shown on the use of the Triplesensor chip for simultaneous measurement of NOx and ammonia.
Development of another NOx sensor technology was reported by Jean-Paul Viricelle [Mines Saint-Étienne]. The sensor operates based on the principle of mixed potential in the Pt/YSZ/Au system. A catalytic filter was added to the sensor in order to oxidize reducing gases (HC, CO), and a third, Pt electrode was added to polarize the Au electrode in order to improve NO2 sensitivity. The sensor is not sensitive to NH3, a NOx accuracy of ±3 ppm at 0 ppm was demonstrated with a prototype of the sensor.
Pressure and Temperature Sensors. A comprehensive overview of pressure sensor technologies and applications was given by A. Graubner and T. Herz [Zentrum Mikroelektronik Dresden AG]. Considering their operating principle, pressure sensors can be divided into (1) passive sensors (i.e., sensors that do not require electricity supply), such as resistive or capacitance sensors, and (2) active sensors, typically piezo-electric. Considering the manufacturing technology, pressure sensors can be divided into (1) ceramic sensors (resistive, capacitance), (2) silicon MEMS sensors (piezoelectric, capacitance), and (3) stainless steel sensors (resistive). Piezoelectric MEMS sensors are most commonly used for exhaust pressure applications, such as DPF monitoring. Stainless steel sensors are used for high pressure measurements, such as in the high pressure circuit of a common rail injection system. In combination with a venturi, pressure sensors can be used for air mass flow measurement, replacing hotfilm MAF sensors when higher accuracy and durability is required—potential applications include heavy-duty Euro VI engines.
New design ideas for self-diagnostics and sensor correction in thermoelectric temperature sensors were discussed by Temperaturmeßtechnik Geraberg [K. Irrgang]. In one novel design, the conventional thermocouple joint was replaced by a sandwiched ring structure that showed a superior dynamic response.
Soot Sensors. A new accumulating, resistive soot sensor for DPF OBD applications has been introduced by Stoneridge [K. Hedayat]. The sensor utilizes a variation of the interdigitated comb structure—known from the commercially available sensor by Bosch—but the digits are very short, while the electrodes are much longer. This design was developed to provide better self-diagnostics capabilities—a fracture in the sensor element prevents current flow through the electrode. To provide fast response, the sensor must collect soot rapidly by utilizing such forces as electrophoresis. This however also attracts more ash into the sensor, which can be a source of error as ash can sinter during sensor regeneration. Another contamination problem is presented by siloxanes added to diesel fuel as antifoaming agents.
Alex Sappok [FST] presented an update on the development of a radio frequency (RF) sensor for direct determination of the soot load accumulated in a DPF. The sensor relies on the measurement of changes in the microwave resonance signal, as the signal is adsorbed by soot. Ash causes a frequency shift in the signal, which also allows to measure ash levels in the DPF. Fast sensor response below 1 s was demonstrated. Performance of the sensor was evaluated over 380,000 mile equivalent DPF aging. In tests on a DDC 13 engine, a reduction in DPF regeneration duration of 15-30% relative to OEM regeneration was demonstrated with the RF sensor.
New Technologies and Applications. Fraunhofer Institute [S. Zische] reported on their development program of microcomponent manufacturing using ceramic thick film and multilayer technology. The application of semiconductor technology in sensors offers such advantages as miniaturization, integration potential, low cost and low power consumption. The long list of sensors that can utilize semiconductor technologies includes pressure, temperature (resistive and thermocouple), flow, soot, gas concentration and humidity sensors.
The new, semiconductor technologies used in automotive sensors call for new methods of analysis for sensor assessment and for detection of faults and problems. Olaf Günnewig [SGS Institut Fresenius] talked about the use of physical surface analysis methods for the characterization of exhaust gas sensors. The discussed methods included electron microscopy (SEM), x-ray diffraction (XRD), x-ray photoelectron spectroscopy (XPS), mass spectrometry (TOF-SIMS), and Auger electron spectroscopy (AES).
University of Bayreuth [Ralf Moos] presented an interesting work on the use of RF sensors as a tool to characterize the operation of emission control catalysts. In the three-way catalyst, the RF sensor was used to determine the state of ceria oxidation—a strong correlation was found between the catalyst oxygen loading and the resonance frequency of the signal. In the SCR catalyst, RF sensors were used to detect ammonia loading. A correlation between resonance frequency and ammonia was found, however, it was dependent on temperature.
OBD Integration. IAV [M. Moser] analyzed the impact of sensor tolerances in the emission aftertreatment system on the robustness of OBD systems. The operation of an SCR based aftertreatment system was simulated using Exothermia software. The impact of five sensor tolerances (NOx upstream, NOx downstream, SCR inlet temperature, mass air flow, DEF dosing) on the output of the OBD system was analyzed. A software tool was developed that allows to determine whether OBD system diagnosis is safe against component tolerances and their combinations, or else if a new diagnostics approach or tighter tolerances are needed.
Matthias Weber [Continental] talked about future trends and evolution of OBD systems. Development trends in powertrain technology impact future OBD systems and sensors, For example, future OBD components must be optimized for reduced exhaust temperatures that are the result of increased powertrain efficiency, and for the reduced exhaust mass flow and pulsation amplitudes in downsized engines. Information technology is another source of a very significant change in OBD systems. Cloud-based diagnostics and services allow for a number of new diagnostics and control functions, including support for shop diagnostics (e.g., augmented reality repair supported by cloud services), fleet diagnostics, statistics and pattern recognition (e.g., component aging), improved warranty/recall tracking, or predictive maintenance.
Critical aspects of cloud-based diagnostic and control systems—not yet properly recognized within the industry—are security and protection of privacy. Vehicles equipped with remote diagnostic systems can transfer data on the vehicle use and other personal information without the driver’s consent or even knowledge. The security threats are also grave, as vehicles—from automobiles to heavy machinery—will have the capability to be remotely controlled and operated over the internet, even in the absence of the driver/operator. Therefore, robust data security and high quality encryption must be used to protect vehicle communication systems and cloud interfaces against hackers. A recent report by US Senator Ed Markey titled Tracking & Hacking: Security & Privacy Gaps Put American Drivers at Risk identified a number of issues with wireless communication technologies in today’s vehicles and concluded: “These findings reveal that there is a clear lack of appropriate security measures to protect drivers against hackers who may be able to take control of a vehicle or against those who may wish to collect and use personal driver information.”