Development of a Novel Multi-Parameter Monitoring Tool for Optimisation and Environmental Compliance of Anaerobic Digestion Technology

Student thesis: Doctoral Thesis


Anaerobic Digestion (AD) technology is used to treat a range of organic feedstocks (e.g. sewages, animal manures, food wastes, industrial organic wastes, and energy crops). It has seen a significant increase in deployment in the last two decades around the world. Due to the wide and sometimes difficult feedstocks range, the process can suffer from instability. Therefore, AD is particularly sensitive to process disturbances by irregular feeding with the potential for microbial toxicity from an accumulation of process intermediates or the presence of inhibitory compounds. This instability can lead to lower conversion efficiencies and reduced biogas production. Another challenge for AD plants is related to the fact that they can be a source of odours, which can lead to complaints from neighbours living close by or close to digestate land spreading areas. Except for gaseous measurements (CH4, CO2, H2S and gas flow), AD plants typically rely on single off-line measurements for solids, total acids and buffering capacity. There is therefore a need to develop monitoring tools that are reliable, that can be used in real-time and that can provide multi-parameter identification and quantification to support plant operations. Besides monitoring strategies can enhance process performance, reduce odour complains, and support environmental compliance and evidence gathering. Here a novel approach using Gas Chromatography-Ion Mobility Spectrometry (GC-IMS) as a multi-parameter monitoring tool to measure volatile organic compounds (VOCs) has been investigated. To enable the development of the tool, a number of samples were sourced from a variety of full-scale AD plants, environmental samples and samples generated from laboratory operated digesters. The majority of samples were sourced from a novel full-scale AD plant that treats food wastes and cattle slurry and was occasionally co-fed with maize silage. The reactor design for this AD plant was based on a plug flow system with gas mixing. The AD plant was chosen due to the diversity of feedstocks, the presence of a homogenisation tank, the type of reactor enabling process intermediates to be sampled, and the existence of a pasteurisation process, and the storage of digestates. Work included the development of an analytical methodology using GC-IMS suitable for AD process related samples. It was the first time that GC-IMS was utilised within this biotechnology industry, except for siloxane measurements in the biogas. Parameters evaluated to optimise peak separations were flows for drift and carrier gas, time for headspace equilibrium, temperatures, and sample preparation (e.g. dilutions, addition of salts). The tool’s performance was evaluated in terms of accuracy, precision, and reliability for identification of several compounds including terpenes, aromatics, ketones, volatile fatty acids (VFAs) and ammonia. The method error was lower than 5% in the repeatability study and 7% in the intermediate precision study. The analytical method was found to be highly robust and the system performed well. In addition to the identification of several key volatiles for AD samples, a preliminary quantification performance was conducted. Initially a 2nd order polynomial equation was used for ammonia hydroxide, limonene, VFAs (acetic, propionic and butyric acids), and several ketones and later a Boltzmann function was used for ketones and limonene. Limits of detection and quantification (LOD and LOQ) were calculated for VFAs using a linear equation, between 0 –500 mg/L.Acetic acid had the highest LOD of 13 mg/Land the LOQ of 40 mg/L. In the case of digestates, which can feature high concentrations of ammonia (in excess of 204mg/L), an impact was felt in terms of carry-over of ammonia. This impact includes loss of the RIP, the creation of extra peaks for terpenes and ketones, and a reduction of detection for aromatics. A solution was established for minimizing ammonia impact, which was based on the addition of an acidifying salt, NaHSO4, which proved to be efficient for concentrations lower than 2.5 g/L of ammonia. For greater ammonia concentrations, a sample dilution would be necessary with the addition of the salt. The tool has enabled the establishment of the fate of volatiles through the AD process(by measuring intermediates and final compounds at full scale as well as a lab scale), and the type and load of odorous compounds within the various samples. In addition, the GC-IMS based spectra analysed by an artificial intelligence based self-organising map (SOM) for classification was found to enable an effective sample comparison identifying the sources of contamination rapidly. With an increase in diversity of samples analysed by GC-IMS associated with a more effective pattern recognition tool, the fingerprinting of environmental contaminants would likely be able to be established rapidly, facilitating the identification of sources of environmental pollutants by regulators enabling the halting accidental spillages. AD plants can have their typical matrices characterised by GC-IMS and in case of a local environmental pollution event, AD plants could protect themselves from being wrongly identified as the source of local incidents that could have been generated by other agricultural or industrial activities and discharges. Finally, this research conducted a preliminary investigation of a laboratory based multi-stage and one-stage reactor setup performance, which concluded that a number of volatiles were reduced in both systems with the multi-stage reactor performing slightly better than the single stage reactors (in particular in the case of terpenes). This study concluded that GC-IMS is a promising analytical tool for multi-parameter diagnosis and control of AD technology. Its ability to analyse a wide range of matrices (gas, liquid and solid), ability to provide rapid measurements with a reduced analytical tool footprint, and the relatively low operator training required can all facilitate its integration as a process analyser for in-situ real-time plant monitoring and control for this industry. Other biotechnologies and biorefining plants could also benefit from the use of this analytical tool (e.g. biochemical and biopolymer factories). Areas of positive impact are expected to relate to feedstock conversions and improved energy/production yields, improved odour management, and environmental compliance.
Date of Award2021
Original languageEnglish
Awarding Institution
  • University of South Wales
SponsorsKESS & Bryn Group
SupervisorSandra Esteves (Supervisor), James Reed (Supervisor) & Tim Patterson (Supervisor)

Cite this