By Ray Hardee
What’s really required is a more careful examination of an entire system to find a solution that meets the business needs of the organization
The expression “less is more” usually doesn’t ring true in our field. Yet, sometimes, as we search for more throughput, capacity and efficiency, adding more pumps or new and more powerful pumps isn’t the answer. It leads to the wrong kind of “more” – more money laid out for capital expenditures, more utility costs, and more disruption and downtime.
“More” is needed, but what’s really required is a more careful examination of an entire system to find a solution that meets the business needs of the organization: to optimize productivity, recover revenue, be more profitable, and – yes – deliver the right level of system performance.
This was the challenge that recently faced a nickel mine and processing mill in Western Canada. It first thought that it needed to add more horsepower and newer and larger pumps to meet increased demand for makeup water. Using simulation software to model the proposed new system, it discovered that the new pumps would drive up electrical costs by 50 percent. In addition, the mine found that the installation process could shut down the mill’s operations for as long as two weeks.
By looking at the entire system, instead of simply focusing on the pumps, the plant staff was able to meet the system’s process requirements without replacing the existing vertical turbine pumps. Shelving the initial plan, the plant engineer determined that by increasing the diameter of the makeup water pump head, and adding a second, wider pipe in parallel to the existing water header, they could meet their objectives for makeup water capacity, at a lower cost in electricity and without a demand charge increase, and without a substantial penalty in downtime for installation and switch-over.
Let’s take a closer look at how the mill and its engineers arrived at this solution.
The mine and mill had been in operation for three decades. Over the years, the makeup water system capacity increased to keep up with expansion. The pump service representative would recommend the addition of impeller stages to the vertical turbine pumps to meet the increased flow and pressure demands. Five years ago, the mill added two pumps to meet the increased flow rates.
Then, during its most recent expansion, the mill called the pump rep again. After looking at the makeup water pumps, he discovered that the desired system flow rate of 4500 gpm could not be achieved using the existing pumps. The rep suggested that the mill replace the existing vertical pumps with five new pumps that were better suited to meet the increase flow rate.
At the outset of the proposed mill expansion, the plant engineer measured a flow of 4170 gpm using an ultrasonic flow meter. The pump rep, asked to recommend what could be done with the installed vertical turbine pumps to increase flow rate, determined that the pumps had reached the maximum number of stages. He suggested running five pumps in parallel to achieve the needed flow rate. During a test, with all pumps running, the measured flow rate was 4600 gpm. Since the plant’s operating procedures require a standby pump for all service systems, the plant engineer needed to add a sixth pump to the system, but there was no room in the pump house for the standby.
At this point, the plant engineer summoned a consultant, who modeled the makeup water system with PIPE-FLO, ESI’s piping simulation program, to better understand what was happening. After modeling the system, he called us, initially to resolve a simple solution that enabled him to match the calculated results to the numbers from plant’s installed discharge pressure gage. Because the results of a simulation are only as accurate as the data entered, we suggested that he apply the manufacturer’s pump performance curve data into the simulation software for superior accuracy.
When PIPE-FLO’s calculated results matched the observed discharge pressure with four pumps in operation, we also suggested that he compare a simulation with five operating pumps to the results of the five-pump test. Once again, the calculated results closely matched the observed plant data. With an accurate representation of the physical piping system in hand, the consultant was able to calculate the pump head requirement.
However, a subsequent simulation for the new pumps revealed the major increase in the electrical load.
The electrical engineer discovered the larger pumps required a 50 percent increase in electrical power consumption. The utility engineer calculated the increase in pumping power would result in a substantial increase in the mill’s demand charges from the electrical utility.
What’s more, the plant engineer projected that the makeup water system would need to be out of service for two weeks to replace the pumps, requiring the installation of temporary pumps to keep the plant in operation. Based upon these unacceptably large expenses and productivity disruptions, the mill’s managers wanted to investigate other options. The consulting engineer sent us his model, and after running the simulation, we quickly discovered the 12-inch discharge header had a fluid velocity of 16 feet per second. The test therefore uncovered another significant problem: a pressure drop of over 500 psi in the main header.
We suggested that the best way to reduce the system electrical load was to increase the pipe diameter of the main header, thereby reducing the system’s pump head requirements. Using this new information, he was able to provide the mill with new options.
The reduction in head loss also resulted in savings of 1618 kw per hour with a savings of $969,144 in pumping cost.
Ultimately, the mine inserted a 20-inch pipeline parallel to the original 12-inch diameter main header. With the two pipelines in parallel, the average fluid velocity in the main headers was 3.5 feet per second, resulting in a pipeline head loss of 52 feet.
The existing 12-inch header with a flow rate of 4500 resulted in a pump head requirement of 1500 feet. Installing a 20-inch header parallel to the existing header results in a pump head requirement of 388 feet of fluid.
The major reduction in head loss allowed the plant to reconfigure the original vertical turbines’ pumps to only five stages. Additionally, only two of the reconfigured pumps were needed to achieve a system flow rate of 4740 gpm.
The reduction in head loss also resulted in savings of 1618 kw per hour with a savings of $969,144 in pumping cost. Additionally, since the facility’s total electrical power consumption went down, there was no increase in the electrical utility’s demand charge. Finally, the new 20-inch diameter parallel makeup water was installed while the existing system remained in operation with only a single day needed to make the tie-in.
Thus, a more system-wide view (rather than a pump-centric one) enabled the mine to:
There was one loose end: The new makeup water system flow requirement was 4500 gpm, but the actual flow rate was 4740 gpm when operating with two reconfigured. This resulted in an excess flow rate of 240 gpm, so what happens to that extra 240 gpm?
After the 4740 gpm is pumped to a reservoir at the top of a hill, the 240 gpm of excess flow returns to the river. This is referred to as by-pass control, and it’s common in many piping systems. In next month’s Pumps & Systems article, we’ll explore how much the bypass control is costing and if there are means of increasing system uptime while reducing costs.