Ultra-low Emission Research Articles

Short Pulse Reductive Activation of Pt/ceria for the Low-Temperature CO Abatement in Vehicles Operated with the Synthetic Diesel Fuel OME

The synthetic Diesel fuel oxymethylene ether (OME) is sulfur-free by nature, and due to the low soot formation, no active filter regeneration events are required, limiting the maximum temperatures seen by the exhaust catalysts to ~ 450 °C. These OME-specific ageing requirements will enable the application of new types of catalysts that cannot be used in conventional Diesel vehicles. Such new catalytic solutions will allow ultra-low emissions at a much-reduced cost and will hence contribute to the overall efficiency of the OME approach. In this contribution, we focus on CO abatement from OME exhaust. To enable an efficient evaluation of new catalysts under practically relevant conditions, a test bench was set up that can reproduce the transient temperature-, mass flow- and concentration profiles measured during real driving tests. In a first step, the transient test bench was used to compare CO oxidation over a commercial Diesel oxidation catalyst for OME- and conventional Diesel conditions. The same low-load cold-start drive cycle run with OME showed slightly lower raw emissions, but the CO emissions downstream of the catalyst increased by a factor of ~ 2. The main reason for the lower CO conversion is the lower temperature of the OME exhaust. In a second step, we investigated short-pulse reductive activation of Pt/ceria as a promising new technology that benefits from the OME-specific low ageing requirements. A Pt/ceria catalyst activated by a short 5–10 s reductive pulse achieved virtually 100% conversion even at exhaust temperatures below 80 °C. With one 5 s reductive activation pulse per 30-minute drive cycle, a CO conversion of > 99.9% is demonstrated over the low-load cold-start OME drive cycle, compared to 59% obtained with a standard commercial Diesel oxidation catalyst. To our knowledge, this is the first published demonstration of short pulse reductive activation of Pt/ceria for CO oxidation using realistic transient drive cycles.

An Offline Closed-Form Optimal Predictive Power Management Strategy for Plug-In Hybrid Electric Vehicles

This article develops an optimal predictive power management strategy (OP-PMS) for plug-in hybrid electric vehicles (PHEVs) to manage the remaining battery level throughout a defined journey. Apart from using the battery before charging it again at the destination, this approach also supports transit through an Ultra Low Emission Zone (ULEZ), where a PHEV needs to operate in pure electric mode to avoid emissions. This type of PHEV power management strategy (PMS) design problem is a noncausal control problem, where the power demand of the remaining driving cycle and the final battery energy level targets influence the current power management action. Rather than solving the whole optimization problem online, this article presents a simplification that leads to a closed-form analytic solution with parameters that can be computed offline. This article makes three key contributions. First, a novel input-linearization method is developed, which converts the OP-PMS into a convex quadratic programming (QP) problem that captures the essential model nonlinearities using an input transformation and a quadratic cost function. Second, the influence of future power demands and final targets on current energy management policy is explicitly quantified from a dynamic optimal control perspective. Third, the noncausal convex QP is approximated with a closed-form analytic solution that is calculated offline. This leads to a real-time implementable OP-PMS with coefficients that can be precalculated and stored in the engine control unit. To demonstrate the viability of this approach, numerical examples are provided based on a generic PHEV model to verify the efficacy of the proposed OP-PMS.

Investigation on low energy-consumed embedded selective catalytic reduction technology for pelletizing flue gas and the CO2 emission reduction assessment.

Based on the unique technique property of grate-kiln pellet production process and its demand of ultra-low emission of NOx, a low energy-consumed embedded selective catalytic reduction (SCR) denitration technology was proposed. The temperature of the NOx-containing tributary flue gas was 350-500 °C, which basically accorded with the temperature range of SCR by V2O5-based catalyst. Considering thepotential inhibition effect of high SO2 concentration (8000-10000 mg/m3) and metal-containing dust in the pelletizing flue gas, the catalyst compositions were optimized to V2O5 (0.5%) and WO3 (5%), giving NO conversion over 90% with low yield of N2O by-product. Compared with the low-medium temperature SCR technology, it was no longer necessary to reheat the flue gas, showinga remarkable CO2 emission reduction effect. The CO2 emission reduction ratios were 94% and 66% contributed by the decrease of fuel and electricity consumption for the embedded SCR technology, respectively. The operating cost was also greatly reduced from 11.4 CNY/t-pellet to 3.1 CNY/t-pellet (Chinese Yuan).

Effects of ultra-low emission air pollution control devices on the evolution of PM and its associated water-soluble ions in a 1000 MW coal-fired power plant

PM (Particulate matter) is a major air pollutant from CFPPs (coal-fired power plants). APCDs (Air pollutant control devices) work in concert to achieve ultra-low emission requirements. Large-scale and higher parameters CFPPs have become a future trend. This work investigated the PM evolution of the APCDs in a 1000 MW ultra-low emission CFPP. The PM1, PM2.5, and PM10 (particulate matter with an aerodynamic diameter less than 1, 2.5, and 10 μm) and WSIs (water-soluble ions) characteristics at the inlet and outlet of selective catalytic reduction (SCR), low-low temperature electrostatic precipitators (LLT-ESP), wet flue gas desulfurization (WFGD), and wet electrostatic precipitator (WESP) were investigated. PM samples were collected by a Dekati low-pressure impactor. The WSIs in PM (K+, Ca2+, Na+, Mg2+, NH4+, Cl-, SO42-, NO3–) were also detected by inductively coupled plasma optical emission spectrometer, ion chromatography, and ultraviolet spectrophotometer, respectively. The results show that PM1, PM2.5, and PM10 generally decrease along with flue gas steam, while PM1 increases by 16.36% and 17.47% at SCR outlet and GGH outlet. The percentages of WSIs in PM show an increasing trend along with flue gas steam. SO42- and Ca2+ are the main components of WSIs. ESP had high efficiencies of 98.18%, 98.54%, and 99.31% for PM1, PM2.5, and PM10, respectively. PM reduces by more than 50% after WFGD and WESP. About 1.00% of PM1, 0.29% of PM2.5, and 0.08% of PM10 are emitted at Stack. PM emissions meet ultra-low emission requirements. The emission factors of PM1, PM2.5, and PM10 are 8.73, 15.90, and 21.04 g/t_fuel on average.

Pandemic restrictions in 2020 highlight the significance of non-road NOx sources in central London

Abstract. Fluxes of nitrogen oxides (NOx=NO+NO2) and carbon dioxide (CO2) were measured using eddy covariance at the British Telecommunications (BT) Tower in central London during the coronavirus pandemic. Comparing fluxes to those measured in 2017 prior to the pandemic restrictions and the introduction of the Ultra-Low Emissions Zone (ULEZ) highlighted a 73 % reduction in NOx emissions between the two periods but only a 20 % reduction in CO2 emissions and a 32 % reduction in traffic load. Use of a footprint model and the London Atmospheric Emissions Inventory (LAEI) identified transport and heat and power generation to be the two dominant sources of NOx and CO2 but with significantly different relative contributions for each species. Application of external constraints on NOx and CO2 emissions allowed the reductions in the different sources to be untangled, identifying that transport NOx emissions had reduced by >73 % since 2017. This was attributed in part to the success of air quality policy in central London but crucially due to the substantial reduction in congestion that resulted from pandemic-reduced mobility. Spatial mapping of the fluxes suggests that central London was dominated by point source heat and power generation emissions during the period of reduced mobility. This will have important implications on future air quality policy for NO2 which, until now, has been primarily focused on the emissions from diesel exhausts.

Numerical optimization of combustion and NOX emission in a retrofitted 600MWe tangentially-fired boiler using lignite

In order to meet the ultralow emission standard of NOx, the combustion system of a 600 MWe tangentially-fired lignite boiler was retrofitted using air-staged combustion technology. The retrofit scheme is described in detail. Combustion and NOx emission characteristics of the original and retrofitted boiler were numerically studied and compared. Optimization simulations for retrofitted boiler were carried out under different over-fired air ratios. The moisture content in lignite was specially considered in simulations by simplified coal compositions method. The results show that the ignition of lignite particles is delayed due to the high moisture content. After retrofit, the temperature level in furnace rises owing to the reduction of total excess air coefficient in furnace, and the high temperature region moves upward due to the introduction of separated over-fire air. The NOx emission is reduced by approximately 40% through retrofitting, which proves that the low-NOx retrofit for the lignite-fired boiler is successful. With over-fired air ratio increasing from 15% to 25%, the NOx emission for retrofitted design decreases from 394 mg/Nm3 to 317 mg/Nm3 (at 6% O2). OFA ratio of 20% is the most economical case, and OFA ratio of 25% is the most environmentally friendly case. Taking NOx emission and boiler thermal efficiency into account, the practical operating condition with OFA ratio of 20% is optimal for retrofitted boiler. The results can be beneficial for the retrofit and practical operation of similar lignite-fired boilers.

Flame Self-interaction and Flow Topology in Turbulent Homogeneous Mixture n-Heptane MILD Combustion

Moderate or Intense Low-oxygen Dilution (MILD) combustion has potential to achieve both high energy efficiency and ultra-low emissions. This analysis adopts the critical point theory to characterise the Flame-Self Interaction (FSI) events and flow topologies in turbulent, homogeneous mixture, n-Heptane MILD combustion using Direct Numerical Simulations (DNS) with reduced chemical mechanism. The local flame geometry has also been categorised using the mean and Gauss curvatures. It was found that the Tunnel Formation (TF) and Tunnel Closure (TC) topologies are the most probable FSI events at all values of the reaction progress variable c, while the Unburned Pocket (UP) and Burned Pocket (BP) topologies were mostly present towards the unburned and burned mixtures of the flame, respectively. Moreover, increasing the turbulence intensity did not result in any significant changes in the distribution of FSI events. Investigation of the flow topology distribution showed that the features associated with non-zero dilatation rate did not exist in the MILD cases considered. This is a consequence of the negligible thermal expansion effect due to the small temperature rise in MILD combustion cases. Increasing the dilution factor caused a reduction in the frequency of FSI events for all c levels. The distributions of flame self-interaction events in homogeneous mixture MILD combustion have been found to be significantly different from previously reported distributions for conventional turbulent premixed combustion.

NOX EMISSION ON RECIPROCAL FILTRATION COMBUSTION OF BIOGAS: A CASE STUDY

Filtration combustion is an effective technique for dealing with low-calorific value fuels, such as derivative fuels from biomass, and simultaneously reducing pollutant emissions. In this study, combustion of biogas (22% of CO2) was investigated theoretically and experimentally by utilizing a reciprocal flow porous burner with heat exchangers inserted in the porous medium. Combustion of lean biogas-air mixtures is stabilized in a naturally transient process, in which a lean equivalence ratio range (0.1 ≤ Φ ≤ 0.9) and gas flow velocities of 0.2 and 0.3 m/s were employed, periodically switching gas flow direction. Reciprocal flow combustion has been compared with unidirectional combustion, using technical methane as reference gas. In this context, predicted temperature profiles inside the burner, as well as experimental results, have shown that the reciprocating system plays an important role in the combustion process, significantly improving flammability limits, efficiency, and emissions. As result, trapezoidal temperature distribution profiles have been obtained with peaks between 1200 and 1600 K, reaching high efficiency up to 90% and ultra-low emission of nitrogen oxides (≤ 2ppm).

Deep purification of low concentration fine particles in a cross flow rotating packed bed

BackgroundThe traditional wet dust removal technology is not effective in removing fine particles with low concentration, which is far from meeting the ultra-low emission and production requirements. However, national emission requirements for fine particles are becoming more stringent year by year. Therefore, it is vital to explore a purification technology for low concentration of fine particles. MethodsCoal-fired fly ash was used to simulate low concentration dust-containing gas and a pilot-scale experimental system was set up employing a cross-flow rotating packed bed (RPB) as the wet separation device to explore the purification capacity of high gravity wet dust removal technology for low concentration fine particles. Significant FindingsThe obtained results suggested that under the operating conditions of inlet dust concentration of 50 mg/m3, high gravity factor of 138.36, gas velocity of 1.71 m/s and liquid spray density of 4.8 m3/(m2·h), the dust removal efficiency was 99.46%. The cut-off particle size of RPB was 0.09 μm, and the fine particles removal efficiency with particle size above 0.2 μm was more than 95%. It can be seen that high gravity wet dust removal technology is a wonderful technology for deep purification of fine particles with strong industrial application potential.

Dynamic emission characteristics and control strategies of air pollutants from motor vehicles in downtown Beijing, China

This study combined the real-time monitoring and investigation of traffic flows to comprehensively analyze the road traffic flow and vehicle structure in downtown Beijing. A dynamic emission inventory of motor vehicle air pollutants in downtown Beijing in 2021 was established, and the impact of these emissions on air quality was simulated and quantified, and different emission reduction control scenarios were proposed to evaluate their environmental improvement effects and explore measures to mitigate the impact of pollution emissions. The results show that the high traffic flow and the structure of the motor vehicle emissions in downtown Beijing are the main causes of severe motor vehicle pollution. Monitoring data shows that traffic flow in central Beijing is dominated by small passenger vehicles, while the vehicle mix is better than in other regions, with 72.0% of vehicles meeting "National V" or higher emission standards. However, to achieve higher air quality goals, further reducing vehicle emissions is necessary. Based on dynamic traffic flow, the average daily emissions of nitrogen oxides (NOX), particulate matters(PM2.5) and volatile organic compounds(VOC) from motor vehicles in central Beijing are 17.7 tons, 0.6 tons and 14.0 tons, respectively, accounting for 23.0% of the city's average daily motor vehicle emissions. If a zero-emission zone for motor vehicles were implemented in central Beijing, the annual average emission level of pollutants will be reduced by 10.4% to 21.0%. The designation of ultra-low emission zones for motor vehicles could be effective in improving the air quality in the center of Beijing.

Variation of nitrate and nitrite in condensable particulate matter from coal-fired power plants under the simulated rapid condensing conditions

In this work, condensation temperature, H2O vapor, SO2, SO3 and NH3 were studied to explore the formation mechanism of nitrate ions (NO3−) and nitrite ions (NO2−) in condensable particulate matter (CPM) discharged by ultra-low emission coal-fired power plants. Some important results were obtained: (i) The concentration of NO3− and NO2− increased with the decrease of condensation temperature, and H2O vapor could also promote the formation of NO3− and NO2−. (ii) The effects of SO2 and SO3 varied at different saturated states of flue gas, which was caused by the redox reaction of SO2 and NOX or the formation of H2SO4. (iii) NH3 could promote the nucleation of NO3− and NO2−, and the promotion effect also existed in the existence of SO2 or SO3. It is worth mentioning that SO3 and SO2 might synergistically inhibit the formation of NO3− and NO2−, regardless of the presence of NH3. The research results would enrich peoples understanding of the chemical and physical characteristics of NO3− and NO2− in CPM and provide a basic reference for the control of CPM emitted from coal-fired power plants.

Real-time prediction of SO2 emission concentration under wide range of variable loads by convolution-LSTM VE-transformer

In this paper, a sulfur dioxide (SO2) emission concentration real-time dynamic prediction model is proposed to achieve the control target of ultra-low SO2 emissions and economical desulfurization system operation at a large-scale coal-fired power plant. First, a vision-expansion self-attention strategy is adopted on the Transformer algorithm structure for better time-series information feature extraction. Second, to adapt to the vision-expansion self-attention strategy, convolution and long short-term memory (LSTM) are combined to process input information position encoding, which can obtain long time-series information absolute position encoding. Then, a Convolution-LSTM VE-Transformer model is constructed by combining the above two proposed parts. The mutual information method and the differential evolution algorithm are combined to calculate correlations between variables and obtain the optimal delay time. Finally, a real-time dynamic prediction model of SO2 is established. In this model, the structure of the input and output variables is determined by the delay time and the influence of coal quality on the SO2 concentration. The quantitative comparison results show the superior accuracy and generalization ability of the proposed prediction model to related popular methods, indicating that the model has strong potential for use in further controller design and applications.

Investigation of syngas combustion in a novel ultra-low emission 20-kW two-stage combustor

Two-stage rich-catalytic lean-swirl combustion is investigated in a 20-kW combustor using biomass-derived syngas as fuel. In the first stage, fuel-rich catalytic combustion of syngas (Фfirst = 4) is investigated using FeCr-alloy monolith and platinum (Pt) as a catalyst. The fuel undergoes partial conversion in the catalytic stage. The remaining unoxidized reactants from the first stage are burnt using additional air in the second stage. The flame is stabilized by two concentric swirling streams in the second stage, where the exhaust from the first stage flows through the inner stream, and the outer stream carries air. By changing the swirl direction of the inner stream, co or counter-swirl flame is established, whereas the overall equivalence ratio is changed by increasing the airflow rate. It is found that the flame topology is significantly different for the co-swirl and counter-swirl configurations. Transitions from a compact ‘M’ shaped flame to a shear-layer stabilized ‘V’-shaped flame is observed by changing the overall equivalence ratio. The combination of planar laser-induced fluorescence (PLIF) and particle image velocimetry (PIV) highlights the differences in the flow field and OH radical distribution, leading to the observed flame shape. The dynamic stability of the combustor is assessed using sound pressure level (decibel) obtained using a microphone and high-speed OH* chemiluminescence. Finally, the CO and NOx concentrations are measured at the combustor exit. The CO emission ranges from 10 ppm to 70 ppm, whereas the NOx emission is negligible in all conditions studied. Thus, the present study successfully demonstrates the feasibility of catalytic and twin-swirl combustion in achieving near-zero emissions while operating with biomass-derived syngas.

A multi-criteria group decision making framework for sustainability evaluation of sintering flue gas treatment technologies in the iron and steel industry

The iron and steel industry (ISI) is at a critical stage of upgrading to ultra-low emission technologies in China. It is imperative to construct a life cycle sustainability evaluation framework as a guide for the selection of ultra-low emission technologies for the ISI. This research developed a novel multi-criteria group decision making framework. Four mature treatment technologies for sintering flue gas (SFG) were selected for analysis from an environmental, economic, and technological multi-dimensional perspective. The results of the life cycle assessment and life cycle costing are incorporated into the evaluation framework. The sustainability of ultra-low emission technologies was evaluated by combining Bayesian and hesitant fuzzy set theories for the first time. The results revealed that of the four alternatives considered, the most sustainable technology consists of semi-dry flue gas desulfurization, bag filter, and selective catalytic reduction denitrification. Triple sensitivity analysis and comparative analysis verified the consistency, robustness, and superiority of the proposed methodology. The interrelationship among criteria showed that technical maturity is key to enhancing the sustainability of ultra-low emission technologies, as it has a significant impact on costs and improves the environmental impact of the system. The conclusions provide optimization suggestions for further improving the sustainability of ultra-low emission technologies for SFG of ISI.

Ultra-low emission standards and corporate production performance: Evidence from Chinese thermal power companies

Finding ways to use coal efficiently and reduce its impact on air pollution is a long-term task that will take sustained work. The impact of ultra-low emission standards (ULES) on corporate productive performance cannot be ignored. Taking the implementation of ULES as a quasi-natural experiment, and using data for listed companies from 2010 to 2018, this paper studies the influence of ULES on the upgrading of thermal power plants using a difference-in-difference (DID) model. The results indicate that ULES significantly inhibits improvements in the productivity of thermal power corporations. This conclusion remained valid after a PSM-DID estimation, the interference of carbon emission trading policy and the 13th Five-Year Plan double control index management on the regression coefficient was eliminated, and a placebo test was used to eliminate interference from other unobservable factors. Further analysis shows that ULES affect corporate productive performance by increasing costs, squeezing out innovation, and reducing efficiency. Moreover, non-state-owned corporations, corporations with highly constrained finances, and corporations in China's eastern region are more obviously affected by ULES. Although ULES for thermal power generation has significantly improved environmental performance, the design of environmental standards should also consider the economic costs, and risks of the green and low-carbon transition process.

A review on research and development of CFB combustion technology in China

The development history and present status of the circulating fluidized bed combustion (CFBC) technology in China are reviewed. This technology has been developed for nearly 40 years since the early 1980s in China, and its development history can be divided into five stages, including a learning stage, an improvement stage, a scale-up stage, a maturing stage, and an ultra-low emission and ultra-supercritical stage. There are currently >3500 units of CFB boilers operating in the country, with unit capacities covering the range of 35–2000 t/h in steam output, mainly used for coal firing. Among the units, the world's first 600 MW supercritical CFB boiler was successfully demonstrated in 2014 and was regarded as a milestone of CFBC technology in the world. CFB boilers are playing an important role in thermal and power generation in China. Several achievements in engineering application and fundamental research are achieved. The reviewed history and achievements could be used as references for researchers and engineers in future CFBC technology development. At the end of the paper, the prospects of CFBC technology in the area of carbon neutralization are discussed. Though the application in coal combustion is shrinking, CFBC technology will play an important role in CO 2 reduction and carbon neutralization. • Present status of circulating fluidized bed combustion (CFBC) technology in China is reported. • CFBC technology development history in China is introduced by five stages. • Major engineering applications of CFBC technology in each stage are reviewed. • Typical fundamental research and achievements on CFBC technology in China are introduced. • The prospects of CFBC technology for carbon neutralization are discussed.

Experiment and simulation studies on thermal performance of a novel water-cooling premixed wall-hung gas boiler

A wall-hung gas boiler was innovatively proposed and designed in this paper. Water-cooling premixed combustion and enhanced condensation heat exchange technology were adopted in the boiler. The extruded aluminum plate-fin structure was adopted in the burner and condensing heat exchanger. Experiment and numerical simulation studies were conducted on the flow, combustion, and heat exchange characteristics of the boiler. The effect of the fin structure, excess air coefficient, heat load, and water-cooling temperature on the thermal performance of the boiler was analyzed. The results showed that reliable ignition, stable combustion, uniform flame distribution, and low pollutant emissions can be achieved in the wall-hung gas boiler. Affected by burning intensity and internal flue gas re-circulation, when the burner plate-fin gap was 1.63 mm, the flame was the shortest and the NOx emissions were the lowest. Under this fin condition, ultra-high efficiency and ultra-low emissions can be achieved in the boiler with a low excess air coefficient. When the excess air coefficient was 1.3, the NOx emissions were less than 30 mg/m3 at the heat load of 4-14 kW, and the thermal efficiency can reach up to 102.8% at the rated load of 14 kW. Within the scope of the experiment, the NOx emissions changed little with the water-cooling temperature. At the temperature of 333 K, there was still a great NOx emission reduction effect.

Analysis on Feasibility of Applying Low-Temperature Emission to Coal-fired Power Plant in Cold Weather

After the ultra-low emission(ULE) transformation, the emissions of ULE coal-fired power plants in China have been reduced greatly. However, considering the total amount of emissions, coal-fired power plants are still among the largest sources. And there may still have great potential in the removal of colored plume and some unconventional pollutants such as condensable particles matter (CPM) and sulfur trioxide (sulfuric acid mist). Though in cold days, hazes occur frequently. Besides, low-temperature condensation is conducive to the condensation and removal of vapor, CPM and sulfur trioxide in the flue gas. In this paper, a new method of low-temperature flue gas emission is proposed for coal-fired power plants in severe cold areas. The flue gas is cooled by ambient air, and discharged without reheating, and remarkable effects can be achieved in removing the pollutants. The technical feasibility and environmental compliance of applying this method to a power plant in cold weather have been analyzed. The results show that low-temperature emission of flue gas for coal-fired power plants in winter is effective in water-saving, emission reduction of pollutants and elimination of wet plume.

Anti-paste Bag Performance of Water-repellent Filter Media in High-humidity Environments Considering Ultra-low Emissions

Anti-paste Bag Performance of Water-repellent Filter Media in High-humidity Environments Considering Ultra-low Emissions

Cylinder Pressure Feedback Control for Ideal Thermodynamic Cycle Tracking: Towards Self-learning Engines

To meet increasingly strict future greenhouse gas and pollutant emission targets, development time and costs of heavy-duty internal combustion engines will reach unacceptable levels. This is mainly due to increased system complexity and need to guarantee robust performance under a wide range of real-world conditions. Cylinder Pressure Based Control is a major contributor to achieve these goals in advanced, highly-efficient engine concepts. Current Cylinder Pressure Based Control approaches use combustion and air-path parameters as feedback signals. These signals are not directly linked to engine efficiency; therefore, compensation for changing ambient conditions, engine ageing or differences in fuel qualities is a non-trivial problem. Contrary to other methods, the method presented in this paper aims to realise an idealised thermodynamic cycle by directly control of the entire cylinder pressure curve. From measured in-cylinder pressure, a new set of feedback signals is derived using principle component decomposition. With these signals, optimal fuel path settings are determined. The potential of this method is demonstrated for a dual fuel Reactivity Controlled Compression Ignition (RCCI) engine, which combines very high efficiency and ultra low nitrogen oxides and particle matter emission. For the studied RCCI engine, it is shown that the newly proposed optimisation method gave the same optimal fuel path settings as existing methods. This is an important step towards self-learning engines.