In this issue, new columnist Harry Rosen from TAS Online and 2KG Training outlines the main differences between the component and systems approaches to pump efficiency analysis and optimisation.
The world is split into two camps when it comes to improving the energy efficiency of pumps. The component-based approach is being driven in Europe through legislation and setting minimum efficiency levels for pump and motor manufacturers. The systems approach has been championed by the USA ever since the US Department of Energy piloted a successful energy savings project in the mid 90s in China. Although it started off as an electric motor optimisation project, it was very quickly discovered that the major savings’ opportunities came from looking at the pumping system, rather than just concentrating on pumps and motors.
The main difference between the component and systems approach comes down to how wide you intend to set your system boundary when evaluating a pumping system. Take the typical pump system of a pump taking fluid from a reservoir and pumping it to a discharge tank a suitable distance away and at a higher elevation – as shown in Figure 1.
The component approach
Let us start with the box surrounding only the pump and motor. Power is measured from the MCC, flow rate from the flow meter situated just downstream of the pump, and head from the difference in pressure between the suction and discharge pressure gauges. The pump efficiency is calculated to be 75.4% and, by comparing this to that on the pump curve from the manufacturer, we find that it is close to the maximum of 79% efficiency for this pump. On a component level the pump is operating efficiently and does not warrant any further attention.
The system approach
Now let us expand the system boundary to incorporate the flow control valve (FCV). This opens to allow bypass flow back to the suction side when demand is low. It is thus not the flow rate through the pump that is important in our example, but the flow to fill the second tank, or to supply a downstream process. The energy consumed for pumping any liquid back to the suction tank is wasted energy.
If we expand the system boundary once again to incorporate the pressure control valve (PRV), we get the true picture of the system demand in terms of flow and pressure. The pressure downstream of the PRV is what the system actually requires, and the pressure loss through the control valve must also be treated as wasted energy.
By using the flow rate at F2 and the pressure after the PRV in our calculations, we can determine the overall system efficiency, which could be dramatically less than our original calculation of pump efficiency. Our system level opportunity would be to remove the throttling valve, close the bypass line and find another way to meet the required system demand – by installing a VSD, trimming the impeller or changing the control methodology, for example.
If we assume that 20% of the flow rate is being returned to the suction tank, and the pressure drop across the throttle valve is around 30% of the upstream pressure, then the overall system efficiency can be calculated to be around 42%. Suddenly there is a major energy savings opportunity. This is the benefit of looking at the system rather than individual components.
A case study: The bypass flow at a sugar mill in the Philippines
The system: Sugar mills provide great opportunities for reducing pumping energy costs. There are numerous pumps used in all aspects of the process, as well as for cooling of process heat. In addition, many mills have cogeneration plants with boilers running on bagasse, the high calorific dry pulpy residue left after the extraction of juice from sugar cane. These plants require additional pumping systems for boiler feed water and cooling pumps for condensing steam back to water.
The system investigated included four hot-well pumps (three operating normally) that pump hot return water from the refinery to a set of spray pans – a low cost alternative to traditional forced-convection cooling towers. The water is cooled down through natural convection by approx 10 °C and then pumped by another set of pumps back to the plant for process cooling.
The level of the water in the sump is below the intakes of the pumps, which can cause ongoing problems for the pumps due to loss of priming and suction lift. Vent pipes had been installed on the discharge of the pumps to facilitate the removal of air from the system when priming the pumps. Plant operators found that if they left these pipes open during normal pumping operation, the stream of water back into the hot well was a visible means of identifying any loss of priming and thus avoiding cavitation and the pumps running dry.
The opportunity: Adopting the systems approach above, these vent lines are functioning similar to a permanently open bypass line, meaning the energy consumed to pump this portion of the flow is wasted. The flow was measured to be in the region of 3.0% of pump flow, which superficially does not seem a lot. However this translates to between 3.0 and 5.0 kW per pump – and in a single pump house operating with three hot-well pumps of 160 kW each and three cold-well pumps of 100 kW each, the total power loss comes to 25 kW, or 197 000 kWh of wasted energy per year. The Philippines currently has the second highest cost of power in Asia (Japan has the highest) and at 10 Pesos per kWh, this ‘waste’ amounts to 1.97-million Pesos (R630 000).
In addition, there were at least another 14 pumps in the plant operating with open-vent pipes, causing total energy wastage of around 410 000 kWh costing P4.1-million (R1.3-milion).
During a pump performance test, we found that the flow rate through Pump #3 was dramatically down, and the pump efficiency was as low as 30%. Even though the vent line was full, the pump had not been primed correctly and had been operating in that condition for more than 24 hrs, wasting excess energy as well as systematically destroying the insides of the pump. This proved that observing a stream of water through the vent line was ineffective in determining whether the pump was primed correctly or not.
Proposed options for rectifying the problem include:
• Install cheap pressure gauges on the suction of each pump to identify any loss of priming as well as the onset of cavitation. By alerting operators to any suction issues and remedying them, energy consumed is reduced and the reliability and life of the pumps is extended.
• Only open the vent (bleed) lines when priming the pump, to get rid of any air in the system. The rest of the time the valves should be closed.
• Use the energy savings from the above project to invest back into more capital-intensive projects, and save even more energy.
In my next column, I will look at the issue of throttling losses through control valves.