Optimization should not make wastewater plants less reliable

2019 ES&E Consultants’ Forum


Patrick Coleman, P.Eng., Black & Veatch

In 1959, R. L. Clark delivered a paper to the Engineering Institute of Canada on the Humber Treatment Plant. In it he stated that:“The pollutional control programme being carried out by Metropolitan Toronto is comparable to any instituted to date on this continent. One of the most important facets of this work is the establishment of a new treatment plant at the mouth of the Humber River.”

At that time, Toronto built a new 227 MLD sewage treatment plant on the Humber Valley Golf Course to serve 475,000 people. The process included raw sewage pre-aeration tanks, primary clarifiers, five aeration tanks, and eight squircular secondary clarifiers. In retrospect, the design of the primary and secondary clarifiers was too aggressive. Engineers later converted the raw sewage pre-aeration tanks to primary clarifiers.

Sixty years later, we are part of a team that is breathing new life into these old aeration tanks. I won’t admit that I think R. L. Clark follows me around the plant, but I do whisper under my breath on occasion: “I read your paper. Yes, we made changes. Yes, your plant is still safe and reliable.”

Subscribe to our Newsletter!

The latest environmental engineering news direct to your inbox. You can unsubscribe at any time.

Thirty years from now, you may be standing in my place. If you are, then remember that your optimization plans should not make the plant less reliable.

What makes a design “reliable”?

A reliable wastewater treatment plant performs its designated function without failure. When parts of the plant fail due to unexpected conditions, the plant should fail safely, protecting public health and plant staff. Flow that enters the plant should always leave the plant without causing flooding or hazardous sewage spills.

When a new design team adjusts capacity to account for changes the original design team did not foresee, the new capacity may be higher or lower than initially envisaged. The team may identify changes that will either unlock capacity delaying an expansion, allow the plant to meet a new stricter effluent limit, or reduce operating costs.

However, if what we are doing makes the plant less reliable and safe, it is not “optimization.” One way to understand this is to consider a person standing a safe distance from a cliff edge. Knowing the danger with wind gusts or unstable ground, they are close enough to see but far enough away that they are safe. Optimization is not moving them closer to the edge so that they can see more. Optimization would be giving them a pair of binoculars. Using capacity provided to operate the plant safely and reliably to accept more flow is moving the plant staff closer to the edge.

What is reliability?

A good designer ensures that a treatment plant reliably meets effluent quality in all but extreme events. Extreme events would be widespread flooding, atypical cold temperatures, extended power outages, natural disasters or malicious vandalism. Extreme events are not historical wet weather or temperature events, or planned maintenance or equipment replacement.

A design achieves reliability by being flexible and robust with some redundancy. Staff can use flexibility to meet new circumstances. For example, a plug flow aeration tank could be designed to switch to a step feed mode during a peak flow event to protect secondary clarifier performance. A robust design performs despite changes in weather or sewage characteristics. For example, a suitably designed activated sludge plant will adjust to load swings in commuter communities between the workweek and the weekend.

However, flexibility and robustness are not enough to achieve reliability. We also need redundancy and excess capacity built into the design that is only required at the edges of the design envelope or for planned maintenance activities. For critical mechanical equipment, this may be a built-in standby. For example, a pump station will always have an extra pump in case one fails. For less critical equipment, the plant may keep parts or a box spare. For example, a thickener may only run for one shift per weekday, leaving enough time for staff to switch out a damaged feed pump.

The decision as to what redundancy unit processes (e.g., aeration tanks) require is not as straightforward as mechanical equipment.

The Water Environment Federation Manual of Practice (MOP) 8 states that the aeration tank design is for controlling the maximum month of the design life at the maximum solids retention time. For this reason, aeration tanks should have “excess capacity” at average loads, especially during the summer months.

In 1974, the US EPA set the minimum design criteria for wastewater treatment plant mechanical, electric, and fluid system and component reliability. The design criteria do not require a redundant aeration tank. However, the rules require that there is at least the equivalent of two aeration tanks. The assumption is that there is enough redundancy in the design that one aeration tank and two clarifiers can treat the sewage when one tank is offline during low flow or warm months.

There is not a single prescriptive answer to unit process redundancy apart from the flow must be able to pass through the plant. Deciding what redundancy is required is an engineering activity done with the operations and maintenance (OPMAN) staff. For example, the expectation for aeration tanks and other unit processes is that they can treat incoming sewage during minor equipment failures, planned maintenance, and all but extreme conditions. The designer must decide with operations how they will make the facility reliable when these events occur.

What is good practice?

“Good practice” is to anticipate events, assess the risk to staff and the plant’s function, and mitigate the risk by incorporating flexibility, choosing robust processes, and including redundant capacity.

Engineers characterize the risk and the associated consequences by answering the following questions:

  • Is this a planned or unplanned event?
  • How long will a unit be unavailable?
  • What else will be offline?
  • Is there a risk of peak flow, high load or cold temperature?
  • What is the consequence of a processing unit being offline?

Designers should check that their design can perform when each of these events occurs. The design should allow for a unit or tank to be taken offline without taking other process units with it. In all cases, the plant requires redundancy for pumping to move flows, to ensure adequate basin and bypass hydraulic capacity, and to back up critical instruments, safety, and DCS/SCADA systems.

When is optimization a danger to staff?

We can become comfortable with risk to the point we cease to recognize the threat that it poses. The disconnection between the consequence and our appreciation of the consequence widens when we do not operate what we design. We are in danger of being swept along with the thinking of others who design but do not operate. We can make the same bad decisions over and over again, edging the design towards a cliff because nothing bad has happened yet. We need to pause and remember that what may be excess capacity for one person may be an insurance policy for another.

We can avoid optimizing out the capacity required to safely operate a plant if we understand the past and current design conditions, understand the risks our decisions place plant staff under, and step through the plant with those that maintain and operate it to learn what their challenges are.

What should be in the optimization report?

If you receive an optimization report that does not include a design basis and an OPMAN risk assessment, then send it back. No one should talk “capacity” without addressing OPMAN issues and understanding all the conditions the plant must perform under. When we re-rate processes, we need to be sure we are not eating away at the “excess capacity” required to operate the plant.

If we don’t follow the question “why is there excess capacity?” with “what would happen if we used it?”, we are in danger of betraying the trust passed onto us to build safe and reliable plants. Don’t you agree, Mr. Clark?

Patrick Coleman, P.Eng., is a Principal Process Engineer at Black & Veatch Canada. This article appears in ES&E Magazine’s December 2019 issue.


Please enter your comment!
Please enter your name here