Market Analysis of WRRF Real-Time, AI-Driven Optimization Solutions

Introduction Water resource recovery facilities (WRRF) face continuing challenges to meeting tighter effluent limits while simultaneously managing or ideally reducing operating costs, energy consumption, and greenhouse gas emissions. To meet strict effluent limits, more complex treatment processes are being installed, which require tight control. These challenges are exacerbated by the loss of experienced operators, with the number of operators projected to decline by 6% by 2032 (Bureau of Labor Statistics, 2023). WRRF artificial intelligence (AI) driven optimization solutions are well positioned to address these challenges, especially to reduce operating expenditures (OPEX) with a focus on energy efficiency, improved biological stability, and even increased plant capacity to some extent. AI-driven solutions can also play a major role in process GHG emissions reduction such as nitrous oxide. Such real-time optimizations relying on Model Predictive Control (MPC) have already provided tangible benefits to WRRFs via different solutions and packages, and as drivers are evolving in the water industry with more focus on 'doing more with less' digital approaches, there will be significant uptake in such solutions deployment by water utilities relying on the latest cloud computing and 'internet of things' (IoT) development. However, the promoted features and benefits can be daunting to understand to build case studies, and there is merit in unpacking such solutions in further detail for the industry's benefit. Therefore, the goal of this paper and presentation is to provide a comprehensive and objective comparison of some of the key WRRF optimization AI-driven optimization solutions currently available accompanied by relevant case studies. This market analysis will benefit water utilities in identifying solutions that best fit their needs and constraints. Each software package was evaluated based on several factors, including: process optimization features, treatment processes optimized, vendor category (SCADA, Water Technology, etc.), optimization software hosting solution and communication architecture, primary outcome such as OPEX reduction, capacity increase, compliance or emissions reduction, and advantages/disadvantages. Each WRRF has its own requirements, so there is no objective 'best' software as utility managers and operators need to select the best software package(s) for their specific circumstances. Methods For each evaluated software package, data was collected and organized from a variety of sources including vendor-supplied literature and websites, public domain technical literature, and the authors' experience. Table 1 summarizes a 'long list' of available optimization solutions that cover both model predictive control as well as advisory dashboards. From this long list, the market analysis then focused on the top seven (7) most prominent vendors with real-world experience with review of case studies and interviews with WRRF managers and operators where specific software packages have been installed. Key criteria used in the evaluation were: 1.Functionality: Evaluating what functions are present in the software, such as the unit processes the software can optimize and the methods used to perform that optimization 2.Performance: Evaluating the effectiveness and robustness of the software in achieving its optimization goals 3.Complexity: Evaluating the degree of difficulty in installing, configuring, and maintaining the software 4.User-friendliness: Evaluating the ease of use and intuitiveness of the software along with available support 5.Openness: Evaluating how easily the software can interface with other systems within a WRRF Results The initial long list of available optimization systems that were considered are summarized in Table 1 and evaluation of the 'top 7' AI-driven short-listed optimization systems in Table 2 with respect to process optimization features and Table 3 with respect to commercialization and solution architectures. Conclusions AI-driven optimization solutions can provide solutions for a number of challenges that WWRFs are facing today and in the near future. However, understanding the functionality, features, and benefits of these solutions can be a daunting task for most utilities. This work has shown the strengths, weaknesses, and suitability of AI-driven optimization solutions for different utility needs. Critically, these solutions are already providing tangible benefits to WRRFs by addressing some of the most pressing challenges facing the wastewater industry. By considering the needs of their individual facility, utility managers can utilize this market survey to assist them in choosing the most appropriate solution for their needs. By making this technology more comprehensible to a wide range of utilities, more WRRFs can take advantage of their benefits, improving the industry as a whole.