Strategic CIP Prioritization Utilizing Optimization Algorithms to Maximize Return on Investment

SUMMARY OF ABSTRACT
Engineers have long been concerned about the risk of corrosion in force mains, and with recent improvements in pressure measurement engineers have become more aware of the risks posed by hydraulic transients. Traditionally, these risks have been considered independently. Recent research indicates that hydraulic transients and corrosion are more interrelated than previously realizeda,b. Western Lake Superior Sanitary Sewer District (WLSSD) and HR Green, Inc. collaborated on this project where the two risks were considered jointly. We will present the results of high-speed hydraulic transient pressure measurement at the headers of three of WLSSD's most critical force mains. We will show how we used calibrated hydraulic transient models to recommend alternatives that considered both of the aforementioned risks. By considering both risks in our recommendations, we have improved the chances that the upgrades will bolster the resiliency of the overall hydraulic system, and not simply trade one risk for another.
BACKGROUND
WLSSD owns and maintains a total of 34 miles of force mains, the majority of which is ductile iron pipe. The utility is concerned about the future longevity of the pipe given that the majority is more than 20 years old, with much of it more than 30 and 40 years old. Due to serving the paper milling industry two of the force mains we will be discussing have relatively high temperature, and high H2S concentrations, putting those mains at greater risk of failure due to corrosion. The hilly nature of the terrain that the force mains traverse creates many opportunities for gas pockets to form at high points, which can exacerbate corrosion risks. In addition to the traditional understanding of the causes of corrosion, recent research indicates that corrosion can also be caused by hydraulic transient pressure cyclesa,b. The high flow, high head, high pump horsepower, and long pipe length also make these lines susceptible to hydraulic transients. The force mains studied in this project are single force mains, without a redundant line, and as a result minimizing the risks to them from corrosion and hydraulic transients is of critical importance to WLSSD.
SCOPE OF PROJECT
In this presentation, we will discuss our findings assessing and addressing the interrelated risks of corrosion and hydraulic transients in three of the larger ductile iron force mains in the state of Minnesota. We will present the results of high-speed transient pressure monitoring, hydraulic transient modeling, as well as the results of our corrosion risk assessment. The three lines studied are the Scanlon Force Main (4.6 miles of 30' and 42' DIP and FRP), the Knowlton Creek Force Main (4.9 miles of 48' and 54' DIP), and the Carlton Force Main (2.2 miles of 14' DIP).
GENERAL PROCEDURES
Hydraulic transients were measured close to the header of each of the three pump stations at a scan rate of 100 samples per second via an Omega Transient Pressure Data Logger (OM-CP-PRTRANSIS-1-500G). An example of the pressure monitoring set up is shown in the attached photo, the data logger is located in the white box on the photo. We measured the force main pressures under typical operating conditions. Understandably, the project team did not want to trigger a high-risk scenario just for the sake of measurement. We built hydraulic models for each pump station and force main that after calibration generally mimicked the results measured in the field. The hydraulic transient models were then used to extrapolate the results across the entire system and to evaluate more extreme events than we were comfortable triggering. The high-risk scenario evaluated for each station was a multi-pump trip (i.e., power failure simulation) of the highest number of pumps that are run simultaneously at each station. The highest number of pumps tripped (in the model) were three, two, and two for Carlton, Scanlon, and Knowlton Creek pump stations, respectively. We performed modeling using KY Pipe Surge Software, which uses the Wave Characteristic Method, and Hazen Williams Equation to calculate the transient pressures.
RESULTS
An example of the measured data overlaid upon the post-calibration model results is shown in the attached Pressure vs. Time plot. An example of a high-risk scenario extrapolation (a 3-pump trip of the Carlton Pump Station) is shown in the attached Pressure vs. Distance plot. From this particular plot we observe that the pressure envelope (in grey), which represents the high and low pressures seen in the pipeline with the recommended number of combination air valves (2) added along the line. As can be seen from the graph, the system with the new valves is exposed to minimum pressures above the level of vapor cavity formation (horizontal pink line) and above -6 psi (horizontal orange line), which provides a sufficient margin of safety against vapor cavity formation. We concluded that to provide a similar level of protection from hydraulic transients seven new combination valves are needed on the Scanlon line, and a single new combination valve is needed on the Knowlton Creek line. The Scanlon line required special consideration because, if all seven combination valves are installed, oxygen (via the air) will be introduced into the force main under normal operating conditions. This will increase the risk of microbiologically influenced corrosion. As a result, we are also considering evaluating other alternatives that do not add oxygen to the line, and therefore do not increase the risk of corrosion. Lastly, a corrosion risk assessment was performed for each force main and we recommend three locations for further pipe integrity testing on each pipe system. We have attached an example of the testing location recommendations for the Carlton force main on Carlton pipe profile.
CONCLUSIONS
- All three pipelines could benefit from improvements to their surge protection systems. - In some cases, existing combination air valves were ineffective due to the existing venting arrangement (or lack thereof). Modeling showed that improving the venting will reduce the observed transients (provided valve slam is not an issue). - In some cases, existing combination air valves did not open under typical or high-risk scenarios, and those valves could be eliminated. - Measurement and modeling of hydraulic transients complement each other, and iterating between the two can provide the most accurate representation of the system in question. - Because surge protection equipment can introduce oxygen, as was seen on the Scanlon force main, it can create an added risk of microbiologically influenced corrosion in wastewater force mains. - Methods of surge protection that do not rely on the addition of air to the system are available and should be considered for problematic pipe profiles in order to address the risks from hydraulic transients without increasing corrosion risks.