Using Mathematical Optimisation in the Water Sector

It's #TechnicalTuesday so Pushpa Vani Murugesan, a Graduate Water Process Engineer in #teamWRc is discussing the benefits of using mathematical optimisation in the water sector!

In simple terms, mathematical optimisation or mathematical programming is choosing the best option out of a set of alternatives, with regard to certain criteria. It is a lot less expensive than building and testing, and a better strategy to use than ‘trial and error’ methods. Today, industries and companies are more driven to make their processes efficient and reduce the consumption of resources. This drive can be attributed to increased awareness about climate change and the need for more sustainable operations.

Here is a practical example of how mathematical optimisation can be applied to the water sector: As part of my Master’s project, I worked on the supply chain optimisation of the digestate, a by-product of anaerobic digestion (AD) which is produced in large quantities and is typically discarded by plant operators. AD is an effective waste-to-energy route for biological waste. However, research on the composition of digestate shows that constituent nutrients like nitrogen make it a suitable fertiliser in the agriculture sector. This provides an opportunity for it to be reused, enabling circular bioeconomy, and replacing traditional mineral fertilizers that are used by farmers.

However, digestate requires treatment before it can be utilised by farmers, to maximise performance. There are many treatment options available. The choice of an appropriate treatment technique needs to consider treatment costs, transportation costs, types of crops in an area, and the policies and regulations in place along with nutrient requirements.

A mathematical model was formulated by taking some of these factors into consideration. Four digestate treatment options were selected for the model; (a) drying, (b) drying and pelletizing, (c) reverse osmosis (RO, nitrification/denitrification and drying) and (d) RO and drying. Data for treatment and transportation costs for the above technologies was obtained from literature. The model was implemented and solved using Pyomo, a Python-based optimisation modelling package. The model aims to act as a decision-making tool for plant operators who might not have the technical know-how to choose the most suitable digestate treatment option.

As resources become scarcer, it is imperative to use existing resources and by-products in the most effective way possible. Tools like mathematical optimisation will pave the way for that.