Rainfall Simulation

Wet weather events generate I/I that enters the sanitary system through foundation drains, defects in the private sanitary laterals, and defects in the public sanitary sewer mains and manholes. Simulated rainfall methods were chosen for this pilot project to achieve on-demand, relatively uniform wet weather conditions for monitoring the flow response without having to wait for actual wet weather events after each round of rehabilitation. Wauwatosa used a combination of dye water flooding and soaker hoses to simulate rainfall. These tests focused on leakage into the laterals but did not evaluate the remaining impact of foundation drains.

Simulated rainfall does not provide the same wet weather conditions as an actual large rain event. For example, the flow into foundation drains is not generated in the simulated rainfall methods. However, the simulated rainfall methods could be implemented quickly after each round of rehabilitation was complete, providing a consistent and repeatable means of simulating wet weather so that the

effectiveness of each round of rehabilitation could be equitably compared. This was especially important during the pilot study because the summer of 2012 was an exceptionally hot, dry year with virtually no wet weather events that were greater than 1-inch. In addition, the antecedent moisture conditions were very dry because of drought conditions and Wauwatosa’s clay soils.

The rate of infiltration into the sanitary sewer laterals was quantified by visually estimating the flow from each lateral using a CCTV camera that was deployed in the sanitary sewer main. In addition, a V-notch weir was installed in a manhole near the downstream end of each of the project areas to measure the cumulative rate of flow during the testing.

3.1.1 Soaker Hose Testing

The soaker hose test is a simulated rainfall method to add moisture to the soil surface in the vicinity of the private laterals. This test attempts to induce infiltration into laterals that is similar to what happens during a moderately large wet weather event. For this test, water from fire hydrants was distributed to perforated hoses that were laid out on each property, on the ground over the location of the laterals.

During the test, water sprayed out of the perforated hoses to wet the ground like a sprinkler. The soaker hose tests were run for at least 2 hours, after which time the sanitary sewer main was televised to observe the lateral flow rates. Dye was not used in the soaker hose tests, so any detected infiltrating water was clear. Soaker hose tests were run continuously for a total of 2 to 8 hours. On average, water was applied at a rate of 10 to 20 gpm per property.

One benefit of the soaker hose test is that all properties are wetted uniformly. The rate of flow to each property was controlled by a valve. Although the flow was not measured at each property, the total flow was measured at the fire hydrants. There was a degree of variability in the flow rate to each property, as well as some variability of hydrant flow rates during each of the rounds of testing. The goal was to deliver approximately equal rates of water to each participating property in all of the testing rounds.

Section 3 Wauwatosa Lateral Rehabilitation Pilot Project

3-2

Use of contents on this sheet is subject to the limitations specified at the end of this document.

Report-WauwatosaLateralRehab_FINAL.docx

Figure 3-1. Soaker Hose Test Schematic

3.1.2 Soaker Hose and Dye Water Testing

After inspecting the lateral flow rates with only soaker hoses in operation, the test was continued by simultaneously flooding the storm sewer with dye water. Plugs were installed in the storm sewer to isolate a block where the storm sewer could be filled with water from a fire hydrant and dye was added to the water from the hydrants. The soaker hose testing continued during the dye water testing. Figure 3-2 shows a schematic of the soaker hose and dye water testing process.

By flooding the storm sewer, additional water was introduced into the soil when dye water exfiltrated from the stormwater system (pipes and catch basins). Dye observed in the CCTV inspection of the lateral flows at the sanitary sewer main indicated a contribution of I/I that originated from the stormwater system. Because the storm sewer was on one side of the street, the laterals that cross under the storm sewer were more vulnerable to I/I originating in the stormwater system. Therefore, higher flow rates were typically observed in the laterals that cross under the storm sewer.

Ideally, the storm sewer was completely filled with dye water so that the catch basins were filled to overflowing; however, in several blocks, it was not possible to surcharge the storm sewer because the ground conditions were so dry that the storm sewer exfiltrated the water at a faster rate than the sewer could be filled. In these instances, the tests were run with the highest possible water level, which was typically after there was no change in the water level in the storm sewer manholes after flooding for a few hours.

Wauwatosa Lateral Rehabilitation Pilot Project Section 3

Figure 3-2. Soaker Hose and Dye Water Test Schematic

3.1.3 Flow Monitoring

The flow in a typical sanitary lateral is a relatively small flow rate that is difficult to measure. Even when there is a relatively large amount of I/I, measurement of flow in a lateral is uncertain. This study used several approaches to quantify the flow in the lateral. These approaches were:

 Visually observing the flow in a lateral from the televising that occurred during the soaker hose and dye water testing;

 Measuring the flow in the sanitary sewer main using a V-notch weir;

 Estimating the flow in the sanitary sewer main based on the water depth in the sewer that was observed during the televising that was completed in conjunction with the soaker hose and dye water testing; and

 For the N. 65^th St. focus area, measuring the flow at a downstream sanitary sewer meter location (FM6 on North Avenue) during testing.

The objective of flow monitoring was to evaluate the change in flow after each phase of rehabilitation.

Several independent means of estimating flow have been used in this study to provide multiple lines of evidence that form the basis for the evaluation. The uncertainty in each of the methods is considerable due to the nature of the flow conditions. Visual observation methods are naturally dependent on the

Section 3 Wauwatosa Lateral Rehabilitation Pilot Project

3-4

Use of contents on this sheet is subject to the limitations specified at the end of this document.

Report-WauwatosaLateralRehab_FINAL.docx

judgment of the inspector. The V-notch weir measurements were used to check the reasonableness of the observed lateral flow values; however, the weir measurements were subject to various uncertainties.

Therefore, the results are based on the collective body of flow estimates from all of the approaches.

3.1.3.1 Direct observation

The flow rate from each lateral entering the sanitary sewer main was estimated from visual observation using the CCTV camera that was deployed in the sanitary sewer main. In this approach, the location of the I/I into the lateral is not known because a lateral launch camera was not pushed up each lateral;

instead the flow rate is estimated at the connection of the lateral into the sanitary sewer main. Flows are estimated during the testing by an inspector that observes the televised flow. The CCTV videos were also reviewed after the test to check the estimated flows for consistency. Because of the subjective nature of the visual observations, the numerical values are not as important as the relative magnitude of the flow.

Larger flow values were assigned to the laterals with the greatest flow.

Figure 3-3 is an example of flow that was observed from a lateral at the connection to the sanitary sewer main. This photo was taken during Round 0 dye water testing on N. 65^th St. in 2010. In this photo, the camera was located inside the sanitary sewer main looking downstream. During televising, the camera stopped a few feet upstream of the lateral connection. Because this property is on the west side of the street, the lateral crosses under the storm sewer that has been flooded with dyed water. A light green tint is visible in the flow. The lateral flow rate is estimated to be approximately 15 gpm.

Figure 3-3. Lateral Flow Entering Sanitary Sewer Main During Dye Water Testing, N. 65^th St., Round 0

Round 0

15 gpm

Wauwatosa Lateral Rehabilitation Pilot Project Section 3

This visual observation method was a quick means of assigning a numerical value to each lateral flow, but the sum total of the flows was supported with other lines of evidence. For example, the accumulation of visual flow estimates was verified against other measurements of flow in the sanitary sewer main near the downstream end of the focus area. The technique was a relatively accurate method of estimating flow.

3.1.3.2 V-notch Weir

A V-notch weir flow meter was installed in the sanitary sewer main near the downstream end of each pilot area. The meter consists of a V-notch weir that was inserted into the sanitary sewer main and a bubbler pressure sensor to detect the water level upstream of the weir.

The V-notch weir monitors the total flow in the sanitary sewer main. This includes the flow from the laterals and as well as other sources of flow such as I/I that directly enters into the sanitary sewer main and manholes, and flow from properties that are upstream of the testing area. As a result, the flow rate measured at the weir should be equal to or greater than the sum of the lateral flow rates.

The 60°V-notch weir that was used for the pilot project provides a very sensitive means of measuring low flow rates. Unfortunately, there were some problems with the meters. The data from most testing days was reasonable, but the data on a few days was inconsistent with other days, and one day the meter failed to operate. The cause of the difficulties is unknown but a couple of possible explanations have been considered. The narrow V-notch may be vulnerable to fouling from solids in the wastewater. If fouling blocked the narrow part of the V-notch, the water level would rise until it reached a level that allowed the flow to pass over the obstruction in the V-notch invert. Because of the sensitivity of the depth-flow relationship, a small change in water depth would convert into a large change in flow.

In some cases, the flow at the V-notch was considerably more than the sum of the observed lateral flows. This may be evidence that the visual estimates of the lateral flows have been underestimated.

Another possible explanation is that dyed water leaking out of the storm sewer may migrate in the storm sewer trench, downstream of the section of pipes that were investigated on any one particular day. Since the V-notch weir was located even farther downstream, the migrating water in the trench may have contributed to I/I into laterals beyond the immediate testing area and contributed to the elevated flow at the V-notch weir. There was a pattern in the V-notch weir data that indicated higher flows when the test area was farther upstream and lower flows at the V-notch weir when the test area was immediately adjacent to the weir.

The notch weir can be a very accurate flow measurement device but in these small focus areas, the V-notch weir results had a significant amount of uncertainty. The data was useful, but the results needed to be interpreted along with other measurement methods.

3.1.3.3 Sanitary Sewer Flow Estimation

Flows in the sanitary sewer main can be estimated from the observed water depth in the CCTV video taken during the soaker hose and dye water testing. As the camera progresses downstream, the

increasing flow in the sanitary sewer main can be observed. The flow rate in the sanitary sewer main can be estimated from the televised images by visually estimating the depth of flow, the percent full level, and the top-width of the flow in sections of the pipe. Depth was estimated for locations that appeared to have uniform flow. Using the estimated water depth, the flow rate was calculated using Manning’s equation which assumes uniform, steady, normal flow conditions and an accurate knowledge of the slope of the pipe. It also depends on an assumed value for the Manning’s roughness coefficient. The incremental change in flow from the upstream to downstream ends of the pipe was compared to the sum of the lateral flow rates. This method was used as another line of evidence along side the other methods that were previously discussed.

Section 3 Wauwatosa Lateral Rehabilitation Pilot Project

3-6

Use of contents on this sheet is subject to the limitations specified at the end of this document.

Report-WauwatosaLateralRehab_FINAL.docx

3.1.3.4 Downstream metering at FM6 on North Avenue

Downstream of the N. 65^th St. focus area, there was an additional flow meter on North Avenue at manhole 317 that monitored flows from April 5, 2012 through December 7, 2012. This flow meter (called FM6 in the previous study of the metershed) monitors the flow from a relatively large area that includes the three block pilot project area on N. 65^th St. The pilot project area is only 16% of the metershed. Data from this meter was reviewed to check for evidence of elevated flows from the simulated rainfall tests in the pilot project area.

All of the approaches have a high degree of uncertainty. By evaluating the flow from all three approaches, the cumulative evidence is useful for drawing conclusions on the effectiveness of the rehabilitation.