This article originally appeared in the October 2021 issue of WQP as "Measure Twice, Cut Once"
It is no secret that pumping eats up by far the greatest quantity of electrical energy for water companies — often some 20% of process operations. With energy-use/energy wastage and the drive towards carbon net zero staying firmly in the spotlight, this makes knowledge of pump performance a truly vital metric.
Historically, the cost of energy has been subsidised by the environment. As the effect of this has been progressively realized, there has been a disproportionately large increase in the cost of energy to consumers. In the U.K., for example, the average household rise in water charges increased by 64% during one recent 10-year period, while at the same time there was an increase of 91% for domestic electrical energy. The industrial rate for electrical energy rose by some 200%. The only option then for a water company to maintain profit is to reduce the unit cost of delivery by increasing efficiency.
Much like a SAT-NAV system works for drivers, pump performance monitoring can be used in conjunction with a decision support system to provide real-time instructions. This can then help guide operators to minimum possible operational costs for a pumping system.
Numerous methods and types of equipment exist to measure pump performance under different circumstances, broadly speaking, in two categories.
The conventional method computes the output power of the pump through measurements of differential pressure across and volumetric flow rate through the pump.
These features include:
- Resilience for all head ranges from zero upwards.
- Requires the measurement of flow rate, therefore; suitable and accurate for a test loop and mostly unsuitable for site use due to absence of straight pipe in the individual pump main.
- Any errors in motor efficiency measurement are propagated directly to an error in pump efficiency computation.
- Costly to implement on site, requiring a flow meter on every pump, which is usually not possible due to physical space constraints, which also limit the accuracy of application.
Second, this method was first documented by a French scientist in 1912. The then named “thermometric technique” used an enthalpy/entropy mapping method to determine pump efficiency without the need for flow rate measurement. The premise is rooted in fundamental thermodynamics in that if you measure any two thermodynamic state variables (TSV) (which include temperature and pressure) then you can calculate any other TSV (which include enthalpy and entropy).
A summary of the features for this method include:
- Purchase cost is lower and irrespective of pump size.
- Installation cost is low.
- Pump efficiency can be measured through a ½-inch BSP tapping on each side of the pump.
- Does not require the straight lengths of the pipe (as needed for the conventional method).
- Suitable for in-situ, on-site pump performance monitoring.
- Accuracy at higher heads.
- Accuracy at low heads less than 33 feet is variable .
- Head limitations to obtain best accuracy.
The outcomes are the same for each method (measurement of head, power, efficiency and flow rate), but each method is suited to different applications.
So, for a 100-year-old technology, why is the thermodynamic method not more widely deployed in the pumping industry? The principal reason has been the requirement to measure very small temperature differentials, which has been much improved of late with the advent of the semiconductor industry. A further reason is not related to the measurement technology, but to the software decision support system, which produces an outcome from the measurement. Great strides have been made in hydraulic analysis methods, and most importantly, how pumping station personnel interact with such a system to overcome the cultural barrier to adoption.
Despite the thermodynamic method being around for more than a century, a breakthrough case study emerged just recently when an international organization operating a water company in the United Kingdom contracted Riventa to install a real-time thermodynamic pump efficiency monitor.
At this large, high-lift station, there are 11 pumps (two of which are not used), ranging in motor-rated power between 1005HP (750kW) and 2145HP (1600kW), with seven units driven by legacy variable speed drives. The head ranges between 262 feet (80m) and 311 feet (95m) and normal flow rates are between 33 million gallons per day and 66 million gallons per day.
A month-long pump monitoring assessment period to understand the current operating regime showed that there were large variations in hydraulic and financial operating behaviour. A 244% variation was observed in the operating cost to deliver 2.2 million gallons for the large pumps (9, 10 & 11).
The pump efficiency of the large pumps varied from 32% to 89%, due to the current operating regime. A 16% variation has been observed in the operating cost to deliver 1 million litres for the small pumps 3, 6 and 7.
The current annual energy bill — before implementation — is estimated to be approximately $5,1 (£3,7M). Therefore, the saving through real-time scheduling is 13.7% or approximately $0.7M (£0.5M) per annum.
A Green Pump Index (GPX) assessment demonstrated that currently, for every $1.4 spent on electrical energy, only $0.66 is used to successfully deliver water to customer’s homes. The rest is losses. With suitable investment in correctly specified best practice technology, it has been calculated that the annual energy cost could be reduced by up to 46%, or $2.38 million through deployment of best practice pump, motor, drive and decision support technology.
A further successful example of the thermodynamic method was demonstrated in Nevada. To help Las Vegas Valley Water District’s (LVVWD) optimize its water distribution, Riventa recently identified potential savings of 12%, equating to $254,000 per year. Pumps were monitored with thermodynamic Freeflow technology at the Hacienda Water Pumping Station (WPS), a strategic part of the district’s network which is also part of the Southern Nevada Water Authority (SNWA).
Previously, it had not been possible to accurately assess the efficiency of the seven fixed-speed drywell pumps which transfer potable water to a regulator tank at another site, but the team soon established that savings of 7.4% would result from refurbishing four of the pumps.
Savings of at least 4.8% may be achieved through pump scheduling at Hacienda WPS, which was built in 1981, with assets upgraded in 1996.
Hacienda WPS is well-designed, with pumps suited to the system, but using Freeflow, the team was still able to identify hidden savings, which are typical for water treatment facilities across the U.S. with equipment such as pumps and blowers of a certain age. Initial data findings showed that the least efficient pump was used the most during the first two-weeks of the project period. Using all data, the team made real-time recommendations for pumping combinations, based on demand, finding savings of 4.8% that could be achieved through pump scheduling. Three pumps required no maintenance, while four were deemed priorities for refurbishment, which would deliver further savings of 7.4%. The technology now provides LVVWD’s operations team with the information they need to operate aging assets and stations as efficiently as possible.
Improvements in technology have the potential to greatly assist with the mandate of guiding operators to the minimum possible operational cost for a pumping system, but suitable selection and deployment can often only be achieved through measurement leading to informed decision making.
As Lord Kelvin rightly stated way back in 1880, “If you can’t measure it, you can’t improve it.”