Investigation of control and on-line optimisation opportunities of a waste-water treatment plant

Northumbrian Water（NW）在泰恩赛德的Howdon工厂厌氧消化多达40，000吨污水污泥（干固体），并处理高达12，000 L/s的原污水。西北400多个污水处理厂产生的所有污泥要么在豪登处理，要么在蒂赛德的类似工厂处理。Howdon对该公司（以及东北地区）的废水处理至关重要。Howdon的高级厌氧消化（AAD）设施旨在通过在热水解工厂（CAMBI）中对污泥进行预处理来实现更高的沼气（生物甲烷和CO2的组合）产量。CAMBI工厂消耗蒸汽形式的能量，以进一步分解AD工艺可用的原料成分，但这样做会产生更多的净能量（即产生更多的沼气），并大大减少AD消化器的停留时间。后来增加了一个沼气“升级”工厂，通过汽提塔及其质量从沼气中去除二氧化碳（热值）提高，使其适合注入天然气网。NW必须管理该工厂，以满足其两个主要目标：即处理所有必要的污泥，使它们能够再循环到陆地上，并优化工厂的收益，同时最小化运营成本。为了实现这一目标，NW必须能够安排计划的维护，适应不可预见的故障，在当地天然气电网容量的限制下工作，决定是否回购自己的天然气，以便为现场发电，并产生和适当利用来自其热电联产厂的余热。但其控制依赖于操作者的经验、可用性和知识。一个监督控制系统，采用当前的数据输入和验证操作选择，将有助于确保污泥处理，优化净收入，允许更多地利用操作员的时间，公司的其他设施，并提供保证，以领导有关工厂的运作。这项工作的目的是调查和制定运营战略（过程控制和优化），以提高对过程的理解、操作和现场稳健性;该项目旨在为废水处理场的运行和控制提供技术解决方案，重点关注豪登场地的运营挑战。这是一个领域，还没有被认为是在以前资助/合作工作之间的纽卡斯尔大学和西北。初步调查已经开发了一个简单的混合线性规划（MILP）模型的一个小面积的网站;工作将进行调整，以适应优化技术的污泥处理网站的其他领域，开发模型，可以在类似的调度方式进行优化。将创建一个全面的多目标回顾性分析器，分析该网站是如何运作的，以及如何运作的。多目标分析仪的目标包括（但不限于）：优化污泥处理量、利润最大化、废物最小化。将对污泥处理量和气体产量之间的关系进行研究，最终将在整个现场优化器中使用，并纳入在线监督控制方案。联机监督控制方案可能直接控制现场，也可能不直接控制现场，但其目的是使用实时操作数据，就现场的可操作性提供准确和最新的建议。

Northumbrian Water (NW) anaerobically digests up to 40,000 tonnes of sewage sludge (dry solids) and processes up to 12,000 L/s of raw sewage at its Howdon site on Tyneside. All sludge arising from NW's over 400 wastewater treatment works are either processed at Howdon or at the similar plant on Teesside. Howdon is critical to the company's (and the North East's) wastewater operations.The Advanced Anaerobic Digestion (AAD) facility at Howdon was designed to achieve higher biogas (combination of biomethane and CO2) volumes by pre-treating the sludge in a thermal hydrolysis plant (CAMBI). The CAMBI plant consumes energy in the form of steam in order to further break down the composition of feedstock available to the AD process, but in doing so producing more net energy (i.e. producing more biogas) and vastly reducing retention times of AD digesters. A later addition added a biogas 'up-grading' plant where CO2 is removed from the biogas via a stripping column and its quality (calorific value) is raised such that it is suitable for injection into the gas grid.NW must manage this plant so that it meets its two primary objectives: that of processing all necessary sludges so that they are capable of being re-cycled to land, and to optimise the revenues accruing to the plant whilst minimising the operating costs. In achieving this NW must be able to schedule planned maintenance, accommodate unforeseen breakdown, work with the constraints of the capacity of the local gas grid, decide whether to buy its own gas back in order to generate electricity for the site and generate and make appropriate use of the waste-heat coming from its combined heat and power plant.Day to day operation of the plant is fully monitored, but its control relies on the experience, availability and knowledge of the operators. A supervisory control system taking current data input and validating operations choices would help in securing sludge processes, optimising net revenues, allowing greater use of the operators time are the company's other facilities, and provide assurance to the leadership regarding plant operation.The aim of this work being to investigate and develop operational strategies (process control and optimisation) in order to improve process understanding, operation and site robustness; the project aims to deliver a technological solution to the operation and control of wastewater treatment site, focussing on the operational challenges at the Howdon site. This is an area that has not been considered in previously funded / collaborative work between Newcastle University and NW.Initial investigations have developed a simple Mixed Integer Linear Programming (MILP) model of a small area of the site; work will be done to adapt the optimisation techniques employed to other areas of the Sludge Processing site, developing models that can be optimised in a similar scheduling fashion. An overall multi-objective retrospective analyser of how the site was operated and how it could have been operated will be created. Objectives for a multi-objective analyser could include (but are not limited to): optimising around Volume of Sludge processed, Maximising Profits, minimising waste.Research into the relationships between volume of sludge processed and gas yield will be undertaken, which will ultimately be used within a whole site optimiser to be incorporated into an on-line supervisory control scheme. The on-line supervisory control scheme may or may not directly control the site, but it is the aim that it will use live operational data to provide accurate and up to date advice on the operability of the site.