Predicting methane yield by linear regression models: A validation study for grassland biomass

Can Biogas Plants Become a Significant Part of the New Polish Energy Deal? Business Opportunities for Poland’s Biogas Industry

The agricultural biogas sector is now facing the opportunity to become a significant actor in the new energy deal as a low-carbon source of electricity. Given the current prospects for rapid growth in the industry, the authors developed an economic model of a medium-sized agricultural biogas plant to assess the rate of return on such an investment. The analysis comprises energy prices, substrates, and other costs reported by the plants already in operation, as well as the electricity sales support system, the actual biogas and electricity yield from the substrates, and the digestate utilisation. It shows that a biogas plant capable of delivering ca. 2000 MWe generates a profit in a much shorter timeframe than 20 years, even under quite uncertain economic conditions. In the model scenario, the breakeven point is reached at slightly below 5000 MWh of power output or at ca. 5800 MWh including financing costs, with a planned annual output of approx. 8000 MWh. The profitability of the model biogas plant was also demonstrated by calculations made for other scenarios which differ in substrate composition and financing structure. The parameters of the econometric model are based on the data collected from a group of 41 units that use only organic plant matter for biogas production.

Solving Regression Problems with Intelligent Machine Learner for Engineering Informatics

Machine learning techniques have been used to develop many regression models to make predictions based on experience and historical data. They might be used singly or in ensembles. Single models are either classification or regression models that use one technique, while ensemble models combine various single models. To construct or find the best model is very complex and time-consuming, so this study develops a new platform, called intelligent Machine Learner (iML), to automatically build popular models and identify the best one. The iML platform is benchmarked with WEKA by analyzing publicly available datasets. After that, four industrial experiments are conducted to evaluate the performance of iML. In all cases, the best models determined by iML are superior to prior studies in terms of accuracy and computation time. Thus, the iML is a powerful and efficient tool for solving regression problems in engineering informatics.

Assessment of a fast method to predict the biochemical methane potential based on biodegradable COD obtained by fractionation respirometric tests

The biochemical methane potential test (BMP) is the most common analytical technique to predict the performance of anaerobic digesters. However, this assay is time-consuming (from 20 to over than 100 days) and consequently impractical when it is necessary to obtain a quick result. Several methods are available for faster BMP prediction but, unfortunately, there is still a lack of a clear alternative. Current aerobic tests underestimate the BMP of substrates since they only detect the easily biodegradable COD. In this context, the potential of COD fractionation respirometric assays, which allow the determination of the particulate slowly biodegradable fraction, was evaluated here as an alternative to early predict the BMP of substrates. Seven different origin waste streams were tested and the anaerobically biodegraded organic matter (COD_met) was compared with the different COD fractions. When considering adapted microorganisms, the appropriate operational conditions and the required biodegradation time, the differences between the COD_met, determined through BMP tests, and the biodegradable COD (COD_b) obtained by respirometry, were not significant (COD_met (57.8026 ± 21.2875) and COD_b (55.6491 ± 21.3417), t (5) = 0.189, p = 0.853). Therefore, results suggest that the BMP of a substrate might be early predicted from its COD_b in only few hours. This methodology was validated by the performance of an inter-laboratory studyconsidering four additional substrates.

Interpretable machine learning for predicting biomethane production in industrial-scale anaerobic co-digestion

The objective of this study is to apply machine learning models to accurately predict daily biomethane production in an industrial-scale co-digestion facility. The methodology involved applying elasticnet, random forest, and extreme gradient boosting to input-output data from an industrial-scale anaerobic co-digestion (ACoD) facility. The models were used to predict biomethane for 1-day, 3-day, 5-day, 10-day, 20-day, 30-day, and 40-day time horizons. These models were fit on four years of operational data. The results showed that elastic net (a model with assumptions of linearity) was clearly outperformed by random forest and extreme gradient boosting (XGBoost), which had out-of-sample R²values ranging between 0.80 and 0.88, depending on the time horizon. In addition, feature importance and partial dependence analysis demonstrated the marginal and interaction effects on biomethane of selected biowaste inputs. For instance, food waste co-digested with percolate were shown to have strong positive interaction effects. One implication of this study is that XGBoost and random forest algorithms applied to industrial-scale ACoD data provide dependable prediction results and may be a useful complement for experimental and mechanistic/theoretical models of anaerobic digestion, especially where detailed substrate characterization is difficult. However, these models have limitations, and suggestions for deriving additional value from these methods are proposed.

Steam explosion pretreatment of rice straw to improve structural carbohydrates anaerobic digestibility for biomethanation

Effectiveness of steam explosion (SE) pretreatment for deconstructing the complex structural carbohydrates (SC) and lignin recalcitrance properties of rice straw (RS) for conjunctive improvement of biofuel yield and waste valorization was evaluated. This work exhibited successful pretreatment of RS at a different pressure (1.2, 1.5, and 1.8 MPa) and retention (3, 6, 9, and 12 min) for enhancement of SC contribution to biomethane production. Regression analysis demonstrated that SE pretreatment efficiency improved at high-temperature and short-retention time for biodegradation of RS. Maximum cumulative methane yield (EMY) achieved 254.8 mL/gvs at 1.2 MPa (3 min) of SE-treated RS with 62.7% of very significant improvement compared with untreated RS (156.6 mL/gvs). Furthermore, solid fraction of xylose, arabinose, cellobiose, glucose, and acid-soluble lignin in SE-treated RS of 1.2 MPa (3 min) were biodegraded by 27.4%, 46.4%, 100%, 48.8%, and 14.1%, respectively, after anaerobic digestion. Therefore, SE pretreatment was an encouraging approach for enhancing SC conversion to biomethane and waste resource to circular economy.