By Benoît Beaudoin and Pierre Michaud
Most municipalities, whatever their size, would benefit from having pumping station monitoring systems that provide valuable information in real time and collect data for performance analysis, including: information on infiltration; overflow volume calculation required by provincial authorities; information to help plan and prioritize maintenance; instantaneous alarms; information for more efficient operation and reduced maintenance costs; and to help pin point intermittent operation problems, and record all the data required for more detailed event diagnostics.
Every day, approximately one-third of pumping stations can have intermittent and abnormal incidents. These include: bouncing starters, faulty check valves, low pumping capacity at a critical time, unreliable level sensors or PLC programming issues.
The cost of a typical real-time pumping station monitoring system can be prohibitive for small- to medium-size communities and often complex to operate. Volucalc, a device from Canadian manufacturer Maid Labs, makes use of very simple information available in the pumping station to collect data and alarm the operator if anything goes wrong. The technology and calculation algorithms used in Volucalc are the result of experimentation in conjunction with the development and manufacturing of thousands of flow meters since 1986.
The system can also be used with more sophisticated control systems to provide additional information that would not be available otherwise. The Volucalc system has already been successfully deployed in a number of municipalities in the province of Quebec and in the United States.
Basics of pumping stations
The US EPA’s Collection Systems Technology Fact Sheet – Sewer Lift Stations states that “power costs account for 85% to 95% of total operational and maintenance cost”. Very few municipalities use this valuable information to evaluate the pumps at their lift stations. With the right tool, getting the appropriate data is simple, inexpensive and can be done in-house. Software can now help analyze data and isolate potential problems and wasted energy.
The most expensive element of pumping stations is the electricity running them. It makes sense to measure the current used by the pumps as the primary way to monitor the operation of the pumps. Most of what is wrong within a pumping station can be tied to the pumps.
If the current from each leg is not properly balanced, chances are that the pump will wear faster and use more electricity to deliver less capacity. If the current is substantially lower or higher on any given leg for all the pumps, then the voltage may be the cause. Measuring the current tells when the pumps start and stop, and, therefore, their run times and number of starts.
The volume differential between the start and stop level measurements allows calculation of flow in and out of the pumping station. If the remote terminal unit (RTU) isn’t properly programmed, if the pumps do not alternate properly, or if starters are chattering, the pumps will be affected, and this will be recorded through the current sensors, among other things.
Everything is important
Four main elements are needed to properly analyze a station: electrical data; integrated smart software; a means to calculate the flow in and out of the wet well; and some basic knowledge and experience with pumping stations.
About one-third of electrical equipment will show intermittent abnormal behaviours at any given time. Being able to measure and analyze these erratic behaviours is the key to predict/report potential problems before they occur. Electrical equipment often gets sick before dying.
SCADA systems monitor the current used by pumps after the inrush peak, but what about the relay and the float switch? If the relay fails to keep the pump on, then the starting float switch might start and stop the pump every few seconds until burning the pump without generating a current based alarm or reaching the high-level alarm. Then, it will burn the other pump. Then, it will reach the high level and the telemetry system will report a high-level alarm.
Every component of a pumping station system is critical. Special analytical functions should be integrated to SCADA systems to detect all abnormal behaviours. Pumping station monitoring tools, like the Volucalc, integrate a number of real-time deductions based diagnostic functions which identify abnormal behaviours before they become critical. It is an inexpensive and easy to install addition to any existing pumping station.
Accurate flow in/out calculations and power consumption by the pumps are mandatory in pumping stations. Each of these elements should be monitored and generate alarms when too high or low. Furthermore, this information is required to calculate the efficiency of the pumps, which is the volume pumped per watt of electricity consumed. This effectively represents the operational cost to get the job done. As an example, a pump that consumes slightly more electricity than before while pumping slightly less than before might not trigger any time or current based alarms but would generate an efficiency alarm from Volucalc.
Control systems are not monitoring devices
The main reason to monitor pumping stations is to minimize the number of catastrophic events and by doing so, reduce the cost of operating and maintaining the system. These two goals are intimately related. Often, maintenance departments react to emergencies when they get an alarm because their existing system does not have the capabilities to offer predictive information. They do not perform the required analysis to be an advanced warning system. This is a missing ingredient common to a lot of systems currently in operation.
Many manufacturers propose their own SCADA solutions and they all do more or less the same thing based on the specifications given by the engineers. If a level gets too high or low, if a current value gets too high or low, if a flow rate gets too high or low, etc., then a warning or an alarm is generated, and an event log is updated in the database. Someone then looks at the reports or screens and decides on the actions to take based on the information. This does not provide for the early detection of abnormal behaviours.
No volumetric flow algorithm programmed in SCADA software has the ability to detect abnormal flow conditions. Understandably, it is a very complex calculation that can be affected by a lot of conditions in the pumping station. Volucalc integrates these capabilities and it takes into account over 100 different variables in doing so.
Should pumps be alternated?
Most pumping stations with constant speed pumps alternate between them, attempting to share the load equally. Is this the most efficient way to operate? Usually, the pump with the lowest flow rate is also the least efficient one. Because of its lower flow rate, it has to pump longer. So, the worst pump wears even faster because it is used more, and its capacity continues to be affected negatively with time. Alternating between pumps is the simplest control solution, but it is not the most efficient one.
A smarter system should detect these efficiency problems and adjust the operation of the pumps in order to use more of the best pump and less of the worst pump instead of the opposite. Until the defective pump is serviced, using the best pump when the flow is high and the worst when the flow is low is a simple solution. By doing so, significant energy saving could translate into a lower electrical bill to operate the pumping stations.
Detecting abnormal conditions
Abnormalities or intermittent issues are not usually catastrophic, and they occur in about 33% of pumping stations without anyone noticing them. These may lead to expensive repairs that could have been prevented if the problem had been handled sooner. The issue is not the capacity to correct but the capacity to detect early.
Even if it happens only once, an advanced warning system should detect when:
- A level sensor transmits a signal that isn’t physically coherent, such as water level falling with no pump in operation or going up and down 10 centimetres within few seconds.
- Sequence of operations of the pumps’ changes, like not alternating at every cycle when they are supposed to do so.
- Capacity of one or all the pumps varies drastically.
- Start and stop levels of the pumps have changed.
- Run time or off time of the pumps does not make sense under normal conditions.
- Improbability in results (e.g., water level falls while pumps are not in operation).
Some of these events are typically non-catastrophic at the time of their first occurrence, but detecting these early means.
- Sending crew to investigate the problem during normal working hours, and when they are available.
- Being able to replace inexpensive components to see if the problem is resolved.
- Not being against the wall with no spare parts in hand or having to procure parts urgently with additional associated costs.
It is a good practice to have a dedicated monitoring device as opposed to a combined control/monitor unit. A defective device often cannot report its own defectiveness, while a standalone monitor could report abnormal behaviours of the controller. When selecting a system, ask about their backup solutions. One should exist.
With advancing technology driving the cost of remote wireless monitoring down, it has become very easy and cost-effective to bring pumping station monitoring to the desk of the operator. Alarm conditions can be texted or emailed, and the operator can visualize the whole pumping station monitoring network at any time. The MaidMaps interface can provide the plant personnel access to real-time data via a web browser.
Each pumping station is shown on a map with instantaneous readings. Recorded data can be graphed and exported for analysis. A number of pre-formatted reports are available to help visualize and monitor the operation of any pumping station on the territory.
The easiest way to lower the operational costs of a collection system is to increase the efficiency of the pumps. Changing the operation sequence to use more of the best and less of the worst pumps is part of the solution.
Smart systems should also have the ability to change the level set points dynamically. As pumps become more efficient as levels go up, operating at a higher set point during a rain event could contribute to additional savings.
The return on investment for the appropriate pumping station monitoring system can be very rapid. Real-time alarming and meaningful monthly reports can help point maintenance personnel in the right direction to avoid emergency situations and costly incidents.
Benoît Beaudoin is with Maid Labs Technologies Inc. Pierre Michaud is with Avensys Solutions Inc. This article appears in ES&E Magazine’s February 2018 issue.