I recently posted a short summary of the 1,800+ page tunnel study prepared by the engineering firm Black & Veatch under contract with Harris County Flood Control District. In today’s post, I will highlight some of the key questions I have about the study.

Before we get to the questions, I wanted to go on record as supporting the use of tunnels to reduce flood risks in Harris County. I support studying them and I support using them to reduce flood risks for areas of the county with the highest current flood risk and the highest social vulnerability.

I also support the use of the January 2021 U.S. Army Corps of Engineers (USACE) Policy Directive known as “Comprehensive Documentation of Benefits in Decision Document.” This policy “emphasizes and expands upon policies and guidance to ensure the USACE decision framework considers, in a comprehensive manner, the total benefits of project alternatives, including equal consideration of economic, environmental and social categories.” I support the use of this approach to help make the case for flood risk reduction projects in areas of the county with the highest current flood risk and the highest social vulnerability.

Now, on to the questions.

Changes to Watershed Screening

Section 3.0 of the report presents the results of a screening process of all 23 watersheds in the county. Black & Veatch screened each watershed to determine if watershed characteristics were favorable for a tunnel system. They considered the number of flooded structures, the social vulnerability of the residents, the location and distance of suitable tunnel discharge points (outfalls), ground elevations needed to allow gravity flow, and the relative cost of tunnels versus traditional flood risk reduction approaches. The table below presents the screening results in summary format.

During their June 16, 2022 presentation, the Harris County Flood Control District (HCFCD) presented a slide that showed the watershed screening results in a map format (below).

While the report indicates that Addicks and Barker Reservoirs were unfavorable for a tunnel system, the map indicated that they both would receive benefits from the installation of a tunnel in those watersheds. Why were these two watersheds presented this way? How can those watersheds receive any benefits from tunnels if those watersheds each have unfavorable characteristics for a tunnel system?

Selection of Evaluation Criteria

As outlined in my prior post, each of the eight tunnel alternatives was evaluated using ten criteria, each with its own weight. The ten criteria were: flood risk, environmental impact, social vulnerability index, operations and maintenance, constructability, community impacts, integration with traditional flood control systems, permitting, land acquisition, and geotechnical. How were the ten criteria selected? Why these ten? Why do some of them include duplicate scoring elements?

Evaluation Criteria Scoring

I have questions about how six of the evaluation criteria were defined and scored.

Environmental Impact: I would expect alternatives with a higher degree of anticipated impacts prior to mitigation would score lower (worse) and those with a lower amount of anticipated impacts prior to mitigation would score higher (better). Was the information in Appendix G – Environmental Constraints Analysis used to estimate the anticipated impacts for each alternative? Why did this criterion include permitting time? Doesn’t that duplicate another evaluation criterion (see below)? Wouldn’t permitting time be similar for all alternatives? Why include potential fines for non-compliance? Wouldn’t HCFCD follow all laws and environmental regulations while implementing any of the alternatives? How was each alternative scored to address the possibility of encountering hazardous materials during construction? Wouldn’t the likelihood of this be the same for all alternatives? Unweighted scores for this criteria range from 12.5 to 17.25. What factors were used to come up with this range of scores?

Operations and Maintenance: All tunnel options are very similar, except for their length and perhaps the number of inlet shafts. The criterion is scored based on the frequency and complexity of operations and maintenance activities. How would these activities differ for each alternative? Unweighted scores for this criteria range from 7.5 to 15.0. What factors were used to come up with this range of scores?

Constructability: All tunnel options are very similar, except for their length and perhaps the number of inlet shafts. The criterion addresses construction risks such as a flood event during construction or constrained surface site access. How would these risks differ for each alternative? Unweighted scores for this criteria range from 15 to 39. What factors were used to come up with this range of scores?

Community Impacts: This criterion includes consideration of the amount of traffic disruption, noise pollution, the number of displaced businesses and homes, and the amount of public outreach “that will be required.” It appears that this criterion considers impacts during project development, construction, and operation. Doesn’t displaced businesses and homes duplicate the Land Acquisition criterion described below? How was the amount of required public outreach scored? Did the report’s authors consider more required outreach a positive or a negative? Most engineers I know consider more outreach a negative, although many planners and elected officials might consider more outreach a positive.

Permitting: This criterion considers “the level of permitting requirements” for each tunnel alternative. This seems like a duplicate of the Environmental Impacts criterion. It appears to be scored higher if less permitting is required and lower if more permitting is required. How would permitting requirements differ for each alternative? Unweighted scores for this criteria range from 8.75 to 18.75. What factors were used to come up with this range of scores?

Land Acquisition: This criterion appears to use at least one of the same scoring elements as Community Impacts. Why duplicate this?

Promotion of Alternatives

As I noted in my summary post, Black & Veatch promoted the Brays Tunnel as a “recommended tunnel concept” because the number of instances of avoided flooding was reduced by assuming that the federal project to continue widening the bayou channel would be constructed in the future. If the federal widening project is not constructed, the construction of a tunnel would increase the number of avoided instances of flooding from 8,700 to 41,252. Why does the report assume that the federal project won’t proceed?

As I noted in my summary post, Black & Veatch promoted the White Oak Tunnel as a “recommended tunnel concept” for reasons that I don’t fully understand. The report states:

The [White Oak Tunnel] rating for flood risk reduction is 1 out of 10, which reflects a relatively low number of instances of structures removed from flooding. However, the cost per instance of structure flooding is considered moderate when compared to all eight (8) tunnel alternatives. Considering this factor and that the alternative received rankings ranging between 4 and 5 for the four evaluation metrics, the White Oak Bayou Tunnel was selected as a “recommended tunnel concept” for further study and refinement.

Why was the White Oak Tunnel selected as a recommended tunnel concept?

Sediment

Black & Veatch considers sediment accumulations in Appendix Q of the report (pp. 1,1818 to 1,849). Heavy storms more easily cause soil erosion and higher runoff flows can more easily move the sediments into our channels, bayous, and any future tunnel systems. This means that sediment will collect in the tunnels and will need to be removed unless we are prepared to accept reduced tunnel capacity as the sediment reduces the effective diameter. In the appendix, Black & Veatch attempts to answer some of the key questions we all should be asking.

Black & Veatch reviewed prior studies and reports sediment concentrations ranging up to 3,180 milligrams of sediment in each liter of water (mg/L). As expected, lower concentrations are associated with lower channel or bayou flow rates and higher concentrations occur with higher flow rates.

This concentration value may not be very meaningful to some of my readers, so I will provide some context. More frequent and smaller rain events generate urban stormwater runoff with sediment concentrations of about 100 mg/L. Imagine pouring some volume of cloudy water with this concentration of sediment floating in it into a collection container. How much water would you need to pour to collect one pound of wet sediment? Using conversion factors we can figure out that you would need to pour about 120,000 gallons of water to collect about 1 pound of sediment. That’s about 5 or 6 days of a garden hose flowing at 15 gallons per minute.

The higher sediment concentration in flood waters would deliver sediment much more quickly. You would only need to pour 4,000 gallons of water with a sediment concentration of 3,000 mg/L to collect about 1 pound of sediment. Since 3,000 mg/L is 30 times larger than 100 mg/L it only takes 1/30th of the flow to get 1 pound collected.

The water flow in the thought experiment above is one way, just into a container. The tunnels will have water flowing in and out at the same time until the floodwaters near the inlets recede. This means the collection of sediment in the tunnel will depend upon other factors, such as sediment particle size (bigger particles settle more easily) and water velocities (higher velocities keep particles from settling).

Black & Veatch notes, in Section 10.3 of the main report, that “sediment will … accumulate in the tunnel” at a rate of “less than a few inches per year.” Black & Veatch suggests that water could be diverted into the tunnel to resuspend the sediment (by moving water over the bed sediment fast enough to exceed the forces that keep it in a nice pile). This water could then be pumped out along with the suspended sediment to keep the tunnel from silting up and losing conveyance capacity over time.

My questions include the following: Where would this water come from? How much water would be needed? What flow rate and volume of water would be needed for each tunnel? To what location would the water and sediment be discharged without harming surface water quality or damaging dredged navigation channels? How much energy would be required to pump the water into and out of the tunnel? What would happen if a large storm occurred while the sediment removal process was underway? How do other tunnel systems handle sediment?

If you think of any questions you’d like to ask, please leave a comment.