Project Case Study: Balanced flow

May 2007 » Feature Articles
Hydraulic modeling software provides an effective design tool to balance wastewater flows to multiple treatment units in a large plant upgrade.
Tiezheng Wang, P.E.

Project
Miami-Dade County, Fla., South District wastewater treatment plant expansion

Civil engineers
Miami-Dade Water and Sewer Department and
Hazen and Sawyer, Hollywood, Fla.

Product application
InfoWorks CS hydraulic modeling software provides an effective design tool to balance wastewater flows to multiple treatment units in a large plant upgrade.


Hydraulic modeling software optimizes treatment plant design in Miami-Dade County.

Florida is the second-fastest growing state in the United States, with statistics showing that 1,000 people a day have moved there during the past two years, particularly to South Florida. Population growth, a consequent construction boom, and a shortage of water resources are the driving forces for optimizing design for a 112.5 million-gallons-per-day (mgd) average daily flow and 450-mgd peak hourly flow expansion for the Miami-Dade Water and Sewer Department’s (WASD) South District wastewater treatment plant in Miami-Dade County, Fla. The plant has a current capacity of 225 mgd, and is one of three wastewater treatment plants in the county of roughly equal capacity, which treat total peak wastewater flows of 680 mgd.

Regional water resources are increasingly stretched, with the key remaining freshwater source being the Everglades—which is essentially a shallow, slow-flowing river discharging to Biscayne Bay and the Florida Straits in South Florida. Regulatory constraints relating to the water shortage mean that all treatment plant effluent standards have to be upgraded to safeguard freshwater sources. County officials considered the following three options for the South District plant:

  • install bigger pipes and pump the wastewater via a 12-mile-long outfall to the Atlantic Ocean;
  • inject effluent via a deep well, 3,000 feet down into a confined groundwater aquifer beneath the aquifer that provides drinking water; or
  • use treated wastewater for irrigation and landscaping.

The county chose the deep-well option, a technique for which Florida’s Department of Environmental Protection (DEP) mandates tertiary treatment to ensure that groundwater is not contaminated by this process. This option predicated a high-level disinfection upgrade for the South District plant.

Additionally, the existing plant had hydraulic restrictions; notably, there were peak flow events exceeding the 225-mgd peak flow capacity. When complete, the South District plant’s 285-mgd, high-level disinfection facility will be one of the largest of its kind in the United States.

Modeling requirements
To design the expansion, and to solve the plant’s problems, a hydraulic model was deemed necessary. Time pressures added by a DEP deadline meant that the project had to be fast-tracked. Initial attempts to address this element of the project using a spreadsheet to calibrate the flows with some additional work in Visual Basic did not work, as essentially guesswork was the only way to estimate how the flows were distributed. Two design teams were involved: the WASD in-house team, designing some portions of the plant; and the Hazen and Sawyer team of engineers, designing different elements of the plant.

The Miami-Dade County Water and Sewer Department’s South District wastewater treatment plant features six, circular settlement tanks.

Among other elements, Hazen and Sawyer’s key task was to design the new filter, and WASD staff designed the chlorine contact tanks. The teams also had to design routing and pipe sizing; design weirs, weir gate openings, and invert levels; and check the flow balance between different units and processes. The correct hydraulic gradient had to be calculated, and the contact tank designed to maintain a 15-minute chlorine contact time.

Hazen and Sawyer decided to use InfoWorks CS software to meet challenges such as defining flows in multiple downstream connections, which involved determining how to distribute flows between four chlorine oxygenation tanks (aeration tanks using pure oxygen for the aeration process) feeding to eight clarifiers. Such situations usually involve creating multiple outlets from the process in question to split the flows, but for a wastewater treatment plant it is also necessary to be able to balance the downstream flows, for which InfoWorks CS was found to work well.

InfoWorks CS also has a real-time control (RTC) capability that can be used to design weir gates, sluice gates, and treatment processes. For instance, one of the features of the plant upgrade is a set of weirs leading into the clarifiers. Modeling such complex features requires a stable hydraulic engine that is unlikely to crash. Other software tasked with this same computation either froze or crashed the computer on which it was being run, but InfoWorks CS handled the challenge well.

The modeling software also provided efficient data input and result interpretation. InfoWorks CS has useful tools such as hydraulic profile layout, reports with drag and drop features, and file and database management for multiple users, who can simply log in and out as required.

A copy-and-paste tool ensured there was no need to repeat work already done. For instance, engineers can simply copy a complex clarifier system design to multiple locations within the plant layout. This feature was useful in other elements of the project, such as designing pump station layouts, where complicated pump systems with maximum and minimum levels, controls, and discharge valves could be copied as a single entity into other locations.

Since extremely flat, rural topography characterizes the county catchment and it is not possible to transport wastewater flows to the plant by gravity, InfoWorks CS could prove useful in modeling the WASD collection/transmission system. Some of the basins within the system contain more than 58,000 manholes and 996 pumping stations, plus as many as 400 private pumping stations, lifting flows to either a force main or another pumping station. The copy-and-paste tool would be a useful and efficient way of creating these elements of the model.

Plant model calibration
The original plant contained four clarifiers and six pure oxygen tanks, as well as other process elements. Sampling points for the calibration process had to be prioritized because there was a limited budget.

Chemicals such as polyelectrolytes are added to the wastewater as it moves through the plant to facilitate the treatment process.

In terms of data collection, there was a need to model peak flow events for the operational calibration points. If no such events occur during the calibration period, it is necessary to create a peak-flow scenario using a forced flow to particular units, measuring the resultant levels, pressures, and flows to enable planning for a storm event. Observations during a storm found some of the clarifiers overflowing as reported, with the overflows dislodging the resident alligator from his home in the detention pond area.

For the historical data, the team consulted plant records and SCADA data. Although the client had 70 flow meters at the plant, not all were working because they are challenging to set up and calibrate. To provide some data, a group of meters was positioned on the base of the clarifier effluent collection trough, with flow depths, rates, velocity, and hydraulic rate measured for comparison against the model.

After the levels were measured, the metered data was found to match well with the findings. Having fixed intervals of measurement made the data easier to interpret. A float tube was designed to provide a water elevation measurement for the oxygenation tanks. Levels are often difficult to interpret because of the lack of light, amount of foam, and turbulence, but the float tube system worked well.

The integrated data was compared against the flow survey information, with all of the metered data set up as one event. For hand-measured data, the time interval was set up to reduce the number of events to be identical to that for the metered data. This meant that the data could all be grouped in one table and plotted on one chart.

When the model was compared against the data from the calibration points, large spikes created during forced-flow events (achieved by shutting valves and channelling increased flows through one flow path) were evident in the metered and hand-derived data, and could be matched in the model by representing closure of the same gates.

Plant expansion
The south District plant expansion involved constructing four new clarifiers, 30 new filters, and chlorine contact tanks, with room reserved for further expansion in the future. The challenge was to channel wastewater through a long line within the existing plant to the new sand filters, and from there back through the chlorine contact tanks, mainly by gravity, with the required flow provided by a transfer pumping station. The client also wanted to run a hydraulic model to help balance flows from the aeration tanks to the new clarifiers.

Running the hydraulic model proved that the suggested changes would work, making it an important tool in the plant’s design process. A workbook was created to summarize alternative modeling scenarios and to assess how much flow needed to go to each unit. Team meetings decided how to balance the flows, run the system, and take the largest processes off line during peak flows as required by the regulators, as well as setting out worst-case scenarios. Once these criteria were decided, they were run through the model to gauge the results.

The team also aimed to achieve better RTC by defining the roles assigned for each control. For instance, the RTC was set to balance the first four, existing chlorine oxygenation tanks simply by gauging the levels required at the downstream weir gates, something the model managed well. The consultants were also able to explain the reason why one of the clarifiers kept overflowing during peak flows—the flow was not well balanced and, as a result, one of the valves kept failing.

In modeling the flow balance for the four existing oxygen tanks, it was necessary to look at the downstream weir gates, which modelled well. The challenge was how to distribute flows evenly to the 30 filter units, particularly since there was a requirement for a 10-percent margin of difference between flows to different flow paths, equating to just 10 mgd of the flow to each unit.

In addition, flows in the pipe system were sometimes propelled under gravity and sometimes by the pressure from the force main. InfoWorks CS enabled engineers to balance the flows so that at any time the flow in one pipe was not greater than the other.

With the chlorine contact tank, the challenge was to maintain hydraulic retention time and minimize head loss. In addition, the groundwater table was high, with rock strata not far below the surface, so digging even an extra unnecessary foot down would have involved a major dewatering program, significantly increasing project costs. Head losses and hydraulic retention times were recalculated to optimize these factors, enabling cost savings on the new filters by making them shallower than the existing chlorination tanks through flow balancing.

Because InfoWorks CS can undertake calculations and export data quickly to provide the desired summary table, a series of trial and error runs was possible, allowing the overall hydraulic gradient to be considered at several critical points.

Worst-case scenarios were considered for various elements of the process, including clarifiers, weirs, and oxygen tanks, ensuring that the filters could be bypassed automatically during peak flows at a set flow height without manually opening valves. The flat topography meant that an accurate hydraulic model was critical, and InfoWorks CS achieved the level of accuracy required.

Conclusion
A reliable hydraulic modeling engine enabled easier network troubleshooting. Other advantages included the following:

  • capability to distribute flows in the gravity-flow conduits;
  • relational database organization and record keeping facilities; and
  • effective interpretation of results and presentation of design operation guidelines.

With a final project construction cost of $405 million, the software solution proved central both to optimizing the design and minimizing the costs. The comprehensive desktop and online help was also a considerable asset.


Tiezheng Wang, P.E., is with Hazen & Sawyer-Florida, based in Hollywood, Fla. He can be contacted at 954-987-0066.


Upcoming Events

See All Upcoming Events