Multiscale Modeling of Hybrid CSTR/Biofilm Based Multi-Vessel Reactors for Anaerobic Digestion Processes

厌氧消化器被用来将农业废物转化为沼气，然后沼气可以燃烧产生电力。最常见的是，这个生物过程是一个1阶段的操作，其中大约50%的生物质能量被转化。更多的能量可能会在第二阶段被提取。对于这种二次工艺，采用固定胶片的方法更为合适。在这个项目中，我们开发了一个模拟工具，能够准确地模拟这样一个两阶段的AD过程。国际水协厌氧消化模型1号(ADM1)是发展厌氧消化过程数学模型的准标准和出发点。对于第一步，这将由高维恒化器类型的模型来建模。为了模拟第二阶段固定膜系统，ADM1可以与Wanner-Gujer类生物膜模型耦合。与耦合这两个模型框架相关的问题是，由于高维，计算需求膨胀，并且模型对于每次运行需要数百或数千个完整模拟的工程任务变得不切实际。其中一个原因是，这种生物膜模型是为了解决生物膜中的浓度场而设计的，而对于在反应器模型中的应用，只需要生物膜和水相之间的底物通量。我们建议开发一个动态的、基于ADM1的厌氧消化生物膜反应器的多尺度模型，该模型正确地解释了生物膜与水相之间的传质，但没有详细地解析生物膜的空间结构，从而提高了计算效率。这样的模型将有助于设计、分析和优化两级厌氧消化器，如本文所讨论的ON。

Anaerobic digesters are used to convert agricultural waste into biogas which then can be burned to produce electricity. Most often this biological process is a 1-stage operation in which approximately 50% of the energy of the biomass is converted. Further energy can possibly be extracted in a second stage. For this secondary process a fixed film approach is more appropriate. In this project we develop a simulation tool capable of accurately modelling such a two-stage AD process. The International Water Association Anaerobic Digestion Model No.1 (ADM1) is the quasi standard and point of departure for mathematical model development of anaerobic digestion processes. For the primary step, this will be modeled by a high-dimensional chemostat type model. To model the second stage fixed film system, ADM1 can be coupled with a Wanner-Gujer-like biofilm model. The problem associated with coupling these two model frameworks is that the computational requirements balloon owing to the high dimensionality, and the models become impractical for engineering tasks that need hundreds or thousands full simulations per run. One reason for this is that such biofilm models are designed to resolve concentration fields in the biofilm, whereas for application in reactor models only the substrate fluxes between the biofilm and the aqueus phase are required. We propose to develop a dynamic, ADM1 based multi-scale model of biofilm reactors for anaerobic digestion that correctly accounts for this mass transfer between biofilm and aqueous phase but do not resolve the spatial structure of the biofilm in detail, thus gaining computational efficiency. Such a model will be useful in the design, analysis, and optimisation of a 2 stage anaerobic digester like the on under consideration here.