Optimal control of combined sewer systems | Daily News


Research at SLIIT

Optimal control of combined sewer systems

Damage to aquatic wildlife in river Thames due to CSOs
Damage to aquatic wildlife in river Thames due to CSOs

Drinking water supply network and wastewater collection network are the two types of water networks available in most of the countries.

The wastewater collection networks gather and transport domestic and industrial wastewater to the wastewater treatment plants. The treatment plant releases treated wastewater to an acceptable quality standard to the receiving water bodies (natural water bodies like lakes and rivers).

However, these wastewater networks are not designed to carry storm water. Nevertheless, during the stormy days, rainwater enters the wastewater networks and mixed with the wastewater. In addition, this creates more volumes inside the wastewater networks. These networks are called combined sewer network in most of the countries. The capacity limitations in the combined sewer networks lead to have combined sewer overflows (CSOs).

These CSOs are identified as one of the major environmental concerns for most of the cities. The untreated CSOs are directly discharged to the nearby natural water bodies and cause many environmental problems including health risks because of the increased pollution levels at natural water bodies.

According to “Utility Week News” on June 8th, 2011, a total of 500,000 tons of storm sewerage were released into the river Thames, United Kingdom, following a heavy rainfall. This resulted in death of a large amount of fish and other aquatic wildlife. This was due to the lowered dissolved oxygen level in the river. Therefore, this is a fine example to illustrate the impacts of CSOs.

Many authorities tried to introduce structural solutions to control the CSOs. These include providing additional storage facilities which are temporary or permanent and to construct more wastewater carrying sewer lines. However, these types of solutions are now outdated due to financial constraints. In addition, the disturbances to the urban inhabits due to various constructions have stepped back the structural measures. Thus, non-structural solutions are peaking for the potential solutions for the environmental damage from CSOs.

Controlling combined sewer networks are not easy, like many other real-world problems. That has many competing and conflicting objectives. Decision makers are struggling to select the best possible control strategy in minimizing the combined sewer overflows (CSOs) when controlling the sewer networks. However, this control strategy should be cost effective to produce a feasible control approach in real world. Cost effectiveness has become significantly important in present economic status of the world. Not so many people would like to spend more on controlling sewer networks. Therefore, the solutions should be more attractive in cost. At the same time, they expect more benefits from the new technology at a lower cost.

Over the past decades, people have witnessed the control strategies based on minimization of CSOs. However, they now want to even ensure the receiving water qualities in rivers and lakes from the CSOs while minimizing the CSOs.

On the other hand, Sri Lanka is no longer a country which relies on controlled or uncontrolled septic tanks as the sewer carried or depositor. Several sewer networks are about to be introduced in various areas in the country (Kandy, Jaffna, Moratuwa, Ratmalana, Ja-Ela, Maharagama and so on) in the coming future. This is basically to protect the groundwater resources of the country from contamination of sewerage. However, sewer lines would be under burden as it is discussed in this article. Therefore, control of combined sewer networks is highly essential to the country too.

Therefore, this research is being carried out to reach a holistic framework that uses multi-objective optimization to control the urban wastewater systems, considering flows and water quality in combined sewers and the cost of wastewater treatment. Along the lines, a comprehensive multi-objective optimization problem is developed to dynamically control the combined sewer networks.

The model minimizes three objectives simultaneously. Cost of wastewater treatment, cost of pump operation in the sewer system and pollution load to the receiving water from CSOs are the three objectives which are considered for the modelling. The modelling process is carried out to as much as match the real-world conditions to the mathematical modelling process.

For example, pollution levels of several water quality parameters in wastewater (domestic and industrial) and storm water runoff are considered. In addition, pollutographs for several water quality parameters are generated for the stormwater runoff. Temporal and spatial variations of the stormwater runoff are incorporated using these pollutographs for different land-uses.

The real world patterns make the control more difficult than many other simplified scenarios. The controlling is not only complex but also it deals with non-linear functions. Therefore, evolutionary algorithms are used to solve this controlling problem. A set of gates are provided to control the sewer system as per the developed multi-objective formulations.

The performance of the multi-objective optimization model is tested on a real waste water system in Liverpool, United Kingdom. Results suggest that the optimization approach can produce dynamic control strategies over the full duration of storm period. Figure 2 illustrates the optimal control settings of gate 1 over the total time period (0-150 minutes) obtained for minimum treatment cost solution.

This clearly shows the dynamic behavior of the gate control based on the multi-objective optimization problem. More importantly, the gate openings at 0-15 minutes time step has influenced the 15-30 minutes gate openings. Gate openings for the previous time step is used to calculate the present and future gate openings.

Therefore, this shows the real time application of the control algorithm. Figure 3 illustrates the gate openings for all gates for the 15-30 minutes time step. These are again based on the developed multi-objective optimization algorithm. The figure clearly shows the dynamic behavior of gate openings and thus the spatial variation of the control setting are showcased. Similarly, Figure 3 presents the temporal variation of the gate openings.

Therefore, the temporal and spatial variation (dynamic control) of the gate openings based on the developed multi-objective optimization problem is feasible. The results demonstrate the benefits of the multi-objective objective optimization approach and its potential to establish the key properties of a range of control strategies through an analysis of the various trade-offs involved. The developed algorithm is not cases pecific, and it is base on a 100% generic algorithm; therefore, the algorithm can be easily adapted to any combined sewer network in world. Further research is carried out to implement this algorithm in the real-time control.

(The research reported here is being carried out by Dr. Upaka Rathnayake, Department of Civil Engineering, SLIIT.)

Visit Kapruka.com Sri Lanka's Largest online shop. Over 125,000 unique categories such as Fresh Flowers, Cakes, Food, Jewllery, Childrens Toys and other Sri Lankan e-commerce categories. Low delivery cost to most cities here and free delivery in Colombo.

Add new comment