The University of Akron. William Troyer The Dr. Gary B. and Pamela S. Williams Honors College

Transcription

1 The University of Akron Honors Research Projects The Dr. Gary B. and Pamela S. Williams Honors College Spring 2018 Applying Control Logic to the End of the Ohio Canal Interceptor Tunnel Based on Downstream Capacity to Reduce Overflow of Akron s Wastewater Collection System William Troyer wrt11@zips.uakron.edu Please take a moment to share how this work helps you through this survey. Your feedback will be important as we plan further development of our repository. Follow this and additional works at: Part of the Civil Engineering Commons, and the Hydraulic Engineering Commons Recommended Citation Troyer, William, "Applying Control Logic to the End of the Ohio Canal Interceptor Tunnel Based on Downstream Capacity to Reduce Overflow of Akron s Wastewater Collection System" (2018). Honors Research Projects This Honors Research Project is brought to you for free and open access by The Dr. Gary B. and Pamela S. Williams Honors College at IdeaExchange@UAkron, the institutional repository of The University of Akron in Akron, Ohio, USA. It has been accepted for inclusion in Honors Research Projects by an authorized administrator of IdeaExchange@UAkron. For more information, please contact mjon@uakron.edu, uapress@uakron.edu.

2 1 Applying Control Logic to the End of the Ohio Canal Interceptor Tunnel Based on Downstream Capacity to Reduce Overflow of Akron s Wastewater Collection System William Troyer ABSTRACT During the Akron Waterways Renewed Program, the city of Akron is studying ways to improve efficiency in the existing sewer system while also reducing the number and scale of combined sewer overflow (CSO) events. Real time control logic can be developed to reduce CSO events by improving flow and storage capacity. Looking at one point in the system, when the target capacity of the downstream pipe was increased from 20% to 80%, the peak volume of overflow from the Ohio Canal Interceptor Tunnel decreased from 9,067 gallons to 5,965 gallons the total system overflow increased from 10.3 million gallons to 28.6 million gallons. By decreasing the target depth of the downstream pipe, thus utilizing the Tunnel for storage, the overall overflow volume was reduced. It was found that when the target capacity of a pipe downstream was increased, the volume of overflow upstream of the pipe decreased, but the total overflow of the system increased. To improve the efficiency of the wastewater system, there needs to be a balance in control logic of multiple control sites based on information from multiple monitoring locations. Key words: Real-time control, combined sewer overflow, control logic

3 2 1. Introduction The city of Akron is currently in the process of updating its wastewater treatment system to comply with an EPA Consent Decree (Akron Waterways Renewed). The project includes many instances of sewer separation of the current combined sewers, green infrastructure, optimization of existing sewers as well as the construction of a large interceptor tunnel under the city called the Ohio Canal Interceptor Tunnel (OCIT). The purpose of this 27-foot diameter tunnel is to provide a place to store stormwater during large rain events when the wastewater treatment facility (WTF) and sewer network may otherwise be overwhelmed. When combined sewer flow is larger than the current WTF, sewer system network and the OCIT can handle, excess combined water is released through overflow outfalls into the Cuyahoga River. CSO events would have detrimental environmental effects from excess levels of pollutants as well as increased erosion from larger flows. These CSO events can also be a financial hardship, as the city would be required to pay fines to the USEPA depending on the size and frequency of the CSO events. A basic schematic of the wastewater collection system is shown in Figure 1. The city is divided into different racks (drainage areas), which contribute storm and sanitary water into the collection system. Figure 1 also identifies the proposed flow monitoring locations. The purpose of monitoring the combined sewer flow is to provide calibration data to update the efficiency of the system. By monitoring the flow at different points in the system, the city can manage wastewater during high rain events in the most efficient way possible. Improving the system will reduce the size and number of overflow events. It will also allow more

4 3 control over the flow into the wastewater treatment plant, causing the plant to run more efficiently and save money. The flow is controlled by several gates and pumps within the system. At these locations, water is diverted into storage basins to be held until the flow downstream decreases to the point where the stored water can be reintroduced. The goal of this research is to develop methods to Figure 1: Schematic of the wastewater collection system for the city of Akron help the existing treatment plant run as efficiently as possible, while reducing the amount of overflow that occurs. This can be accomplished by a real time control system (RTC) and a real time decision support system (RT-DSS). A real time control system consists of monitoring and logic and is completely automated. Water is diverted to and from storage based on real time measurements taken at key locations determined through simulations. A real time decision support system

5 4 builds on real time control by attempting to predict future flows based on current conditions as well as predictions such as weather forecasts. RT-DSS is not automated, but rather assists human decision-making. The benefit of RT-DSS is that it is much more nuanced and helps operate the system through prediction, rather than just reacting to current conditions. This project will look specifically at beginning to develop real time control logic at the end of the OCIT, located near Hickory St. in northwest Akron, and assessing which logic has the best impact on reducing the system overflow. The effect of the logic used will be to turn on and off a pump to send water into storage basins at the end of the OCIT or allow it to continue to flow through to the WTF, based on the capacity of the pipe downstream. This paper will present the results of the impact on controlling the water storage based on the capacity of the inputting the wastewater treatment facility and controlling with three different capacities. 2. Methods a. Modeling The model was initially developed in ICM by Mott McDonald and converted into PySWMM by EmNet. The model conversion into PySWMM was necessary to perform advanced analysis of control logic development. The model consisted of an input file that modeled the structures and how the

6 5 system performed, and a model file that produced rainfall data. This project looked at storm over a 24-hour period (April 12, 1994) that occurred in the typical storm data from the model file. Running the simulations in python made it easier to introduce logic to change how the different structures in the system operate. Python language also allows for coding that produces different types of output files which produce several results of attributes at any location. This allows the results to be Figure 2: Display from PCSWMM model denoting pump, storage, and monitoring locations. OCIT is the Ohio Canal Interceptor Tunnel, US is the upstream pipe, and DS is the downstream pipe. The wastewater treatment plant is located downstream of the shown area. easily organized and presented. b. Site Selection and Location This project will look at controlling one pump based on the capacity flowing through one pipe. Figure 2 shows the location of the key points that are either monitored or where logic is applied. The pump will be controlled based on the capacity of the downstream pipe labeled DS. The locations and characteristics that will be monitored are: total volume of overflow from the OCIT tunnels measured at the outfalls total volume of overflow in the entire system total volume stored in the 3 storage basins at the end of the OCIT the capacity in pipes at four locations: the OCIT before the storage (labeled OCIT ), the pipe upstream of the node where the OCIT joins the rest of the system (labeled US ), the pipe

7 6 downstream of the same node (labeled DS ), and the capacity of the entering the wastewater treatment plant (labeled WWTP) The volume of the total overflow of the system is produced in a report file after the simulation. These values were also compared. The logic applied opened or closed the pump to keep the capacity of the downstream pipe at the target value. The target values used were 80% (assumed to be approximately full capacity), 50%, and 20%. The results could also be compared with the default logic. The default control logic opened and closed the pump based on a target depth of 2.6 feet (40% capacity) in the upstream pipe. The results were given in tables showing attributes at each timestamp as well as overall system total values. 3. Results and Discussion The following figures were developed from the tables that were the output of the simulations. The tables used are in Appendix A. Figures 3 through 6 shows how the overflow at the end of the OCIT compares with the OCIT storage. It helps visualize that overflow at that location only occurs when the storage has reached the maximum capacity. The storage basins begin to fill up when the pump is turned on. Therefore, there is more storage and overflow when the target capacity of the downstream pipe is lowered.

8 7 Figure 3: Volumes of overflow and storage occurring at the end of the OCIT for the default control simulation. Data from Table 1 Figure 4: Volumes of overflow and storage occurring at the end of the OCIT for 80% downstream capacity target control simulation. Data from Table 3

9 8 Figure 5: Volumes of overflow and storage occurring at the end of the OCIT for 50% downstream capacity target control simulation. Data from Table 5 Figure 6: Volumes of overflow and storage occurring at the end of the OCIT for 20% downstream capacity target control simulation. Data from Table 7

10 9 Figure 7: Comparison of the amount of overflow occurring at the end of the OCIT with each control logic simulation. Data from Tables 1, 3, 5, and 7 The amount of overflow that occurs at the OCIT for each target capacity is shown in Figure 7. Here, it is easy to see that when the target capacity is reduced, more overflow occurs at the OCIT. It also compares the amount of OCIT overflow that occurs from the default controls. The OCIT overflow from the default controls is very similar to the amount of overflow occurring when the downstream pipe has a target capacity of 20%. For both the default logic and the 20% capacity logic, the peak amount of overflow at a given time is just under 10,000 gallons. When the target capacity of the downstream pipe is set to 80% the peak overflow is reduced to 6,000 gallons. Figure 7 also demonstrates how it takes longer for overflow to begin and an overflow event ends sooner when the capacity is increased. At 80% target capacity overflow occurs from 8:05 AM to 9:35 AM compared to 5:30 AM to 12:20 PM for the target capacity of 20%.

11 10 Figure 8: Comparison of the amount of storage at the end of the OCIT with each control logic simulation of target values of 80%, 50%, and 20% capacity as well as the default control logic. Data from Tables 1, 3, 5, and 7. Since the OCIT storage is related to the OCIT overflow, Figure 8 is similar to Figure 7. However, it is helpful to see how much water is in the storage at the end of the simulation. Since the simulation starts with everything empty, the volumes of each control start at zero. However, Figure 8 shows how, at the end of the simulation, there is water being stored when the capacity is controlled at a downstream target capacity of 20%. This means that if another storm would occur shortly after the first one, the storage would fill up faster and create more overflow.

12 11 Figure 9: Capacity of the inputting the wastewater treatment plant comparing each method of control logic. Data from Tables 1, 3, 5, and 7. Figure 9 compares each simulation of control logic with the capacity of inputting the wastewater treatment plant. The simulation begins with the empty and each simulation fills the at approximately the same rate. The reaches full capacity (80%) at about the same for all the simulations but remains at full capacity longer for the 80% and 50% target capacities. For the default controls and the 20% target, the capacity of the decreases at about the same rate. However, at about 12:24 PM, the default control simulation experiences an increase in flow. While the capacity of the for the 20% target simulation begins to decrease before the 80% and 50% simulations, it decreases at a slower rate. At the end of the measured period the 20% target ends with a capacity at 46%, while the 80% and 50% target capacities end with the 40% full.

13 12 Figure 10: Capacity of the pipe downstream of the OCIT branch. This chart compares the result of each method of control logic applied. Data from Tables 1, 3, 5, and 7. Figure 10 is like the previous figure, in that it compares the capacities in the downstream pipe. This is the location where the control logic was applied. This means, except for the default controls, the pump is operating to maintain the downstream pipe at the desired capacity. However, the pump only controls OCIT so any excess capacity comes from the upstream pipe. It is likely that, when the capacity is above the target value, the OCIT flow is being stored, and the downstream flow comes from the upstream pipe. It also shows a drastic fluctuation when the control capacity is set to 50%. This was caused by the pump turning on and off due to the capacity being below 50% until water is released from storage, bringing the capacity back above 50%. The downstream pipe is at full capacity (80%) for less time than the into the wastewater treatment facility. Again, the 20% target simulation is at a lower capacity for the most of the time period

14 until the 80% and 50% target simulations experience a steep decrease and end with a lower pipe capacity then the 20% target simulation. 13 Figure 11: Capacity of the pipe upstream of the OCIT branch. This chart compares the result of each method of control logic applied. Data from Tables 1, 3, 5, and 7. Figure 11 compares the capacities in the upstream pipe for each target capacity. When the capacity target in the downstream pipe is larger, the capacity also increases. The capacity of the OCIT follows very similar trends to that of the downstream pipe (Figure 10). However, the capacity in the OCIT is slightly larger than that of the downstream pipe. For example the 20% target simulation ends with the OCIT at 31% versus 27% for the downstream pipe.

15 14 Figure 12: Capacity of the Ohio Canal Interceptor Tunnel. This chart compares the result of each method of control logic applied. Data from Tables 1, 3, 5, and 7. Figure 12 compares the capacity in the OCIT for each control logic. It has similar results to the previous figures. As the target capacity downstream is increased, the capacity in the OCIT also increases. The upstream pipe capacities follow a very similar trend to that of the other pipes monitored in the previous figures. The flow does take up less capacity in the upstream pipe then it does in the other pipes. While the OCIT and downstream pipe approach 80% capacity for the 80% and 50% target capacities, the upstream only briefly exceeds 50% capacities for the same simulations (occurring at 4:48 AM).

16 15 Figure 13: Total overflow for the entire system for each method of control logic applied. Data from Tables 2, 4, 6, and 8. The total overflow in the entire system for each simulation is shown in Figure 13. This figure shows that by decreasing the target capacity in the downstream pipe, the total overflow also decreases. Compared with the default controls, setting the downstream target capacity to 20% results in a similar total overflow (10.5 MGD versus 10.3 MGD). 4. Conclusions This project only focused on one point in the entire wastewater system for the city of Akron. However, the results can be used to help move forward to develop better controls for operating the system more efficiently. Three main conclusions were drawn from the results. The first conclusion is that when the downstream pipe is operating near maximum capacity (80%), there is a reduction in overflow from the OCIT. In other words, when a pipe downstream of a pump or gate is targeted to operate at high capacity, there will be less overflow occurring upstream of the gate

17 and pump. This makes sense as it means the gate or pump is allowing more water to head downstream rather than sending it to storage or an outfall. 16 The second result found was that, as the target capacity of the downstream pipe was increased, the total system overflow also increased. Referring to the first conclusion made, this means the system-wide overflow increased, despite the OCIT overflow decreasing. It can be concluded that there needs to be a balance between control sites. When the target capacity was high for the downstream pipe, it allowed a higher flow of water to enter from the OCIT. This left less room in the pipe for flow coming from the upstream pipe, thus causing more overflow upstream of the upstream pipe. These instances of overflow had a larger influence than the overflow of the OCIT. In order to further reduce the total system overflow, a balance needs to be found between multiple control sites. A third result found was that, as the target capacity in the downstream pipe decreased, the total system overflow also decreased. The conclusion is that utilizing the storage at the end of the OCIT helps reduce the total overflow. However, this result may change based on the length, duration, and number of storms in a given time period. To better develop logic that can reduce the overflow of the wastewater system for the city of Akron, more research should be done. Instead of using control logic based on capacity downstream, branches can be dewatered depending on either available storage or the flow upstream. Also, the control logic should be tested using different types of storm events to improve its efficiency. Finally, different sites can be tested and monitored simultaneously to find the balance between each area of the system.

18 17 5. Appendix A: Data Tables Table 1: Results from default controls simulation. sim current time Overflow_Transition XXXtoXXXOverflowX XXCulvert.1 Overflow_Transition XXXtoXXXOverflowX XXCulvert.2 combined overflow ocit culvert 1 overflow volume ocit culvert 2 overflow volume OCIT Overflow Volume Volume_Transition XXXtoXXXOverflow XXXCulvert Volume_Over flowxxxweir XXXStorage Volume_Tunnel XXXDiversionXX XStructure OCIT Storage Volume plant depth plant flow 12:05:04 AM % % % % 12:10:06 AM % % % % 12:15:07 AM % % % % 12:20:10 AM % % % % 12:25:11 AM % % % % 12:30:13 AM % % % % 12:35:13 AM % % % % 12:40:14 AM % % % % 12:45:15 AM % % % % 12:50:15 AM % % % % 12:55:16 AM % % % % 1:00:17 AM , % % % % 1:05:18 AM , % % % % 1:10:19 AM , % % % % 1:15:20 AM , % % % % 1:20:22 AM , % % % % 1:25:25 AM , % % % % 1:30:27 AM , % % % % 1:35:30 AM , % % % % 1:40:33 AM , % % % % 1:45:34 AM , % % % % 1:50:34 AM , % % % % 1:55:35 AM , % % % % 2:00:36 AM , % % % % 2:05:38 AM , % % % % 2:10:39 AM , % % % % 2:15:40 AM , % % % % 2:20:40 AM , % % % % 2:25:41 AM , % % % % 2:30:42 AM , % % % % 2:35:42 AM , % % % % 2:40:42 AM , % % % % 2:45:42 AM , % % % % 2:50:43 AM , % % % % 2:55:43 AM , % % % % 3:00:44 AM , % % % % 3:05:44 AM , % % % % 3:10:44 AM , % % % % 3:15:45 AM , % % % % 3:20:45 AM , % % % % 3:25:46 AM , % % % % 3:30:46 AM , % % % % 3:35:46 AM , % % % % 3:40:46 AM , % % % % 3:45:46 AM , % % % % 3:50:47 AM , % % % % 3:55:47 AM , % % % % 4:00:47 AM , % % % % 4:05:48 AM , % % % % 4:10:48 AM , % % % % 4:15:48 AM , % % % % 4:20:48 AM , % % % % 4:25:48 AM , % % % % 4:30:48 AM , % % % % 4:35:48 AM , % % % % 4:40:48 AM , % % % % 4:45:48 AM , % % % % 4:50:48 AM , % % % % 4:55:48 AM , % % % % 5:00:48 AM , % % % % 5:05:48 AM , % % % % 5:10:48 AM , % % % % 5:15:48 AM , % % % % 5:20:48 AM , % % % % 5:25:48 AM , % % % % 5:30:48 AM , % % % % 5:35:48 AM , % % % % 5:40:48 AM , % % % % 5:45:48 AM , % % % % 5:50:48 AM , % % % % 5:55:48 AM , % % % % 6:00:48 AM , % % % % 6:05:48 AM , % % % % 6:10:48 AM , % % % % 6:15:48 AM , % % % % 6:20:48 AM , % % % % 6:25:48 AM , % % % % 6:30:48 AM , % % % % 6:35:48 AM , % % % % 6:40:48 AM , % % % % 6:45:48 AM , % % % % 6:50:48 AM , % % % % 6:55:49 AM , % % % % 7:00:49 AM , % % % % 7:05:49 AM , % % % % 7:10:50 AM , % % % % 7:15:50 AM , % % % % 7:20:50 AM , % % % % 7:25:50 AM , % % % % 7:30:50 AM , % % % % 7:35:50 AM , % % % % 7:40:50 AM , % % % % 7:45:50 AM , % % % % 7:50:50 AM , % % % % 7:55:50 AM , % % % % plant capacity downstream depth downstream flow downstream capacity upstream depth upstream flow upstream capacity control branch depth control branch flow control branch capacity

19 18 Table 1 (cont.): Results from default controls simulation. 8:00:50 AM , % % % % 8:05:50 AM , % % % % 8:10:50 AM , % % % % 8:15:50 AM , % % % % 8:20:50 AM , % % % % 8:25:50 AM , % % % % 8:30:50 AM , % % % % 8:35:50 AM , % % % % 8:40:50 AM , % % % % 8:45:51 AM , % % % % 8:50:51 AM , % % % % 8:55:51 AM , % % % % 9:00:51 AM , % % % % 9:05:52 AM , % % % % 9:10:53 AM , % % % % 9:15:53 AM , % % % % 9:20:54 AM , % % % % 9:25:54 AM , % % % % 9:30:54 AM , % % % % 9:35:54 AM , % % % % 9:40:55 AM , % % % % 9:45:55 AM , % % % % 9:50:55 AM , % % % % 9:55:56 AM , % % % % 10:00:56 AM , % % % % 10:05:57 AM , % % % % 10:10:57 AM , % % % % 10:15:58 AM , % % % % 10:20:58 AM , % % % % 10:25:59 AM , % % % % 10:30:59 AM , % % % % 10:36:00 AM , % % % % 10:41:00 AM , % % % % 10:46:00 AM , % % % % 10:51:01 AM , % % % % 10:56:01 AM , % % % % 11:01:01 AM , % % % % 11:06:02 AM , % % % % 11:11:02 AM , % % % % 11:16:03 AM , % % % % 11:21:04 AM , % % % % 11:26:05 AM , % % % % 11:31:05 AM , % % % % 11:36:05 AM , % % % % 11:41:05 AM , % % % % 11:46:05 AM , % % % % 11:51:06 AM , % % % % 11:56:07 AM , % % % % 12:01:07 PM , % % % % 12:06:07 PM , % % % % 12:11:07 PM , % % % % 12:16:08 PM , % % % % 12:21:08 PM , % % % % 12:26:08 PM , % % % % 12:31:09 PM , % % % % 12:36:09 PM , % % % % 12:41:09 PM , % % % % 12:46:09 PM , % % % % 12:51:11 PM , % % % % 12:56:12 PM , % % % % 1:01:13 PM , % % % % 1:06:15 PM , % % % % 1:11:16 PM , % % % % 1:16:17 PM , % % % % 1:21:19 PM , % % % % 1:26:20 PM , % % % % 1:31:22 PM , % % % % 1:36:23 PM , % % % % 1:41:24 PM , % % % % 1:46:25 PM , % % % % 1:51:27 PM , % % % % 1:56:27 PM , % % % % 2:01:29 PM , % % % % 2:06:29 PM 8.99E E , % % % % 2:11:30 PM 8.01E E , % % % % 2:16:30 PM 7.18E E , % % % % 2:21:31 PM 6.47E E , % % % % 2:26:31 PM 5.86E E , % % % % 2:31:33 PM 5.32E E , % % % % 2:36:34 PM 4.85E E , % % % % 2:41:35 PM 4.45E E , % % % % 2:46:37 PM 4.19E E , % % % % 2:51:38 PM 3.98E E , % % % % 2:56:39 PM 3.78E E , % % % % 3:01:39 PM 3.60E E , % % % % 3:06:41 PM 3.44E E , % % % % 3:11:42 PM 3.28E E , % % % % 3:16:43 PM 3.14E E , % % % % 3:21:44 PM 3.01E E , % % % % 3:26:46 PM 2.89E E , % % % % 3:31:47 PM 2.77E E , % % % % 3:36:48 PM 2.66E E , % % % % 3:41:50 PM 2.56E E , % % % % 3:46:50 PM 2.46E E , % % % % 3:51:51 PM 2.37E E , % % % % 3:56:52 PM 2.28E E , % % % %