Houston Wastewater Utility Perspective: Balanced O&M and CIP Budget with Projected Outlook of Risk and Level of Service is Now a Reality

Objectives & status: The City of Houston Public Works (HPW) has a sewer network that has a long-standing history of experiencing Sanitary Sewer Overflows (SSO/SSOs). A majority of the Public SSOs are caused by Fats, Oils, and Greases (FOG) that accumulate in the sewer system, restricting the flow rate of the pipe and causing upstream assets to overflow. It is estimated that roughly 70% of these SSOs would be preventable (Image 1) if the City focuses its O&M and CIP resources towards the key factors (or root-causes) of these dry weather SSOs. In the past years, to combat this issue HPW performed a mostly reactive approach, deploying cleaning and inspection resources when a service request was generated after SSOs were notified.
The objective of this project is threefold:
1. Quantify asset-level risk of SSOs for the entire city with non-traditional data science techniques.
2. Identify root causes of SSOs and their individual effect in each asset's risk level.
3. Ultimately find the balance between O&M and CIP Budget, as well as the right initiatives to minimize the risk of SSOs under the consent decree's requirements and in a proactive manner.
The City of Houston started this project in 2020 with the support of GoAigua, which provided the data analysis and software platform for the generation of insights. After piloting the methodology in a group of 13 lift station sewersheds, the methodology has been scaled to the full city and will be rolled out soon.
Methodology/ approach: The project's unique methodology combined a traditional asset management approach with machine learning techniques for the quantification of risk and root-cause analysis of SSOs and FOGs generation: First, HPW and GoAigua developed a methodology to quantify the asset-level risk and consequence of SSOs. For this methodology HPW tested different machine learning algorithms, and eventually selected a random-forest technique. This allowed them to identify all root causes ultimately affecting the generation of SSOs and quantify their effect on the risk to SSO. Some of the key factors include presence of multifamily units, grease generator establishments (e.g., restaurants), average flow velocity, material of pipes, etc. (See Image 2) Once all factors were assessed at an asset, sewer basin and City level, HPW could evaluate potential risk mitigation strategies to minimize the occurrence of SSOs – both O&M initiatives, and CIP ones.
Finally, all these initiatives were piloted and evaluated individually in three selected sewersheds (see Image 3) with a high level of SSOs, and different root causes affecting them. The evaluation was done along two key dimensions: cost/ investment, and benefit/efficacy, which allowed us to prioritize and select them for a larger deployment at a City level. Key initiatives included: public awareness campaigns for grease generator establishments, increased control over grease traps, increased preventative maintenance/ cleaning in risky assets (higher frequency), investment in IoT sensors and real-time surveillance. Findings and significance: Our presentation showcases how the City of Houston was able to leverage their AIAP platform along with machine learning techniques with near real-time and other data sets, both form the Public Works Department and from other Departments (e.g., Health Department, Flood Control, etc.) This project will provide an environmental and economic benefit to the City, as well as improving the customer experience. This allows HPW to better utilize the limited resources, and have full visibility of the balance between O&M vs. CIP budgets.