Proactively Evaluate Sedimentation Risks in the Design of Inverted Siphon of Sanitary Sewer System, A TRA Case study

This paper presents a sedimentation risk evaluation via modeling for a sanitary sewer siphon system design project in Carrollton, Texas and Farmers Branch, Texas, suburbs of Dallas, Texas. The audience will learn about selecting dry and wet weather hydraulic conditions to use for evaluating sediment, how to use modeling to proactively identify the predicted locations and severity of sedimentation risks to assist designers and utilities to select design alternatives, predict sedimentation accumulation, and incorporate sedimentation control and maintenance strategies during the design phases. The Trinity River Authority (TRA) of Texas is the largest water and wastewater service wholesale provider in Texas. TRA is in the process of developing rehabilitation and replacement plans for portions of its Central Regional Wastewater System (CRWS) Elm Fork Interceptor. The project involves the design of three channel crossings. For the crossing at Cooks Branch (Figures 1 and 2 ) in the City of Carrollton, the proposed crossing is an inverted siphon (hereafter, 'siphon') with parallel 42-inch and 66-inch pipes (Figures 3 and 4). There is a proposed weir upstream of the 66-inch barrel, so it only conveys the wet weather flow (WWF), and the dry weather flow (DWF) will only pass through the 42-inch pipe. The siphon is sized to convey projected 2070 future peak WWF during a 5-year design storm. Since the DWF is much lower than the WWF, the designers and TRA had concerns about sedimentation issues in the siphon at low-flow velocities. TRA has a hydrologic and hydraulic model in Autodesk InfoWorks ICM. The model can perform DWF and WWF projections and hydraulic grade line simulations of the collection systems for existing and design conditions. It was proposed to use ICM's water quality module to analyze sediment fate and transport for the siphon system. The Ackers-White method is chosen for the Erosion / Deposition Model. ICM can simulate two sediment fractions independently, but only one fraction (SF1) was used for this analysis, assuming the suspended solids particle size compositions and settling characteristics have no significant variations that warrant refined differentiations. The solids size is based on the average sediment particle size parameter of d50. Based on literature references, average and high sediment parameter values for particle sizes d50, concentrations, specific gravity and settling velocity were used for the evaluation (Table 1). The high sediment parameters represent high risks of sedimentation accumulation for the worst conditions. Long-term historical rainfalls were used for WWF projection as it occurred in the area and are expected to repeat in the future. Historical annual rainfall data (1899-2023) from the Dallas/Fort Worth (DFW) International Airport weather station was collected from NOAA's National Weather Service website and statistically analyzed (Figure 5). The lowest annual rainfall is 17.91 inches in 1921. The median rainfall is 33.59 inches in 2021. Unfortunately, finer time step (15-minute) rainfall data necessary for modeling is not available for the DFW station. However, the 15-minute time step rainfall data for recent years is available at a nearby Fort Worth WSFO Station. The data from 2000-2013 (Table 2) shows the 2008 rainfall of 12.4 inches is less than the 1921 rainfall, and the 2009 rainfall of 34.5 inches is close to the median rainfall of 2021. The two calendar years of rainfall contain a very dry year-duration with higher risks of sedimentation accumulation followed by an average year with average wash-out condition. Therefore, it is considered a reasonable rainfall period to evaluate the sedimentation risks for the siphon. A 5-year design storm was added at the end of 2009 to evaluate additional cleaning effects at the design condition. The simulated flow to the siphon is shown in Figure 6. The peak flow is 78.6 mgd for the 5-year design storm. The following summarize the results (Table 3) and conclusions of the evaluation: Average sediment parameter inputs: - No sediment issue is predicted for the 42-inch barrel. - For the 66-inch barrel: --Upstream Junction: The sediment accumulation at the weir location is due to stagnant water. --The influent pipe to the Downstream Junction: It has incremental sediment build up during dry weather periods due to low recirculation flow from the 42-inch pipe. The downstream end can be washed out by most of the storms. The upstream end has accumulations. High sediment parameter inputs: - The 42-inch barrel: --It can have 3.5 hours of sediment accumulation during early morning low-flow period, but the sediment will be washed out during higher flow hours (Figure 7). - The 66-inch barrel: --Upstream Junction: Sediment accumulation occurs for the first two sections downstream of the weir due to stagnant water (Figures 8 and 9). --The influent pipe to the Downstream Junction (Figure 10):It has incremental sediment build up during dry weather periods due to low recirculation flow from the 42-inch pipe. The downstream end can be washed out by most of the storms. The upstream end has accumulations. --The rest of the pipes can self-clean (Figure 11) - The sediment doesn't increase the head loss significantly for the 5-year storm after a 2008-2009 rainfall condition. However, it could be further verified by detailed modeling such as computational fluid dynamics (CFD) modeling. Based on the evaluation, the following were recommended: - The current Cooks Branch Siphon sizing is confirmed and recommended. - A Texas Commission on Environmental Quality (TCEQ) variance may be required as the weekday peak DWF velocity is less than 3.0 ft/s. - Consider the following sediment management options for the 66-inch pipe: --A back flow prevention device. --An inspection program and maintenance plan to remove the accumulation periodically. --Supplemental water for sediment cleansing. --Sluice gate at 42-inch pipe inlet. Conclusion/Recommendation: The sediment accumulation in collection systems can be modeled in conjunction with Hydrologic and Hydraulic models with available modeling software. The model uses typical sediment and erosion/deposition theories and equations. Reasonable sediment parameter ranges and DWF/WWF projections are required for the modeling. Using average and more conservative conditions for both sediment parameters and rainfalls can provide sensitivity analysis of sediment risks in the systems. The model can identify potential sediment accumulation locations/causes and help refine the designs to minimize the impact by incorporating operation and maintenance strategies into the design. Status of Completion: At the time of submittal, the study confirms the design and sizing of the Cooks Branch Siphon. Based on the sediment location/cause results, the design team is considering several options to manage the potential sediments. Similar analysis will be applied to two additional channel crossing sites.