Straightforward hydraulic systems follow simple rules. For non-Newtonian and incompressible fluids, flow and net pressure drop are directly related. Consider the water transfer system shown in Figure 1. Wastewater generated in Unit 1 is stored in tank T1. Some of the water is reused in Unit 2. The remaining wastewater goes to treatment via vessel V1 in Unit 2. When Unit 2 isn’t running, centrifugal pump P1 provides the head for transferring the water from tank T1 to vessel V1.
Plant management wants to increase the average water rate by 20%. Equipment elevations must remain the same, as must upstream or downstream process conditions. Because the system has simple hydraulics, providing more flow requires some combination of reduced head losses in the system or increased pressure generated by pump P1.
The control valves balance the pump performance curve against the system curve. As long as the control valves can open and the flow rate is within the pump’s capability, more flow is possible.
In this case, with the control valves wide open, the new flow rate can’t be achieved. Even with modifications the system is in a gray area. Some simple modifications may — or may not — allow the desired flow rate.
Already identified modifications include:
• moving exchangers E1/E2 into parallel; and
• replacing orifice flow elements FE1 and FE2.
Plant management must accept some combination of:
• lower flow;
• removal of more pressure drop; or
• increase in head available.
The solution may include all or any combination of these steps. Let’s look at the flow scheme and then examine each area in turn.
The system has a fixed static head loss. Because the equipment and process don’t change, the static losses don’t either. The entire pressure-drop reduction must come from cuts in dynamic head losses. Three components — the exchangers E1/2, the control elements and the piping — mainly contribute to the dynamic head losses.
Exchangers E1/2 originally were in series. At the higher flow rate, the units will operate in parallel. They are multi-tube exchangers with the inlet water on the tube side. Unless they are completely replaced, the exchangers now have a minimum pressure drop.
The control elements include two orifice flow elements plus two control valves. The modifications already include replacing both flow elements. Changing the orifice plates saves 13.2 psi of pressure drop. At the desired rate, the two new flow elements will have a combined pressure drop of 5.1 psi.
After the modifications are made and with the existing pump, if it performs per the manufacturer’s performance curve, the system can meet the required flow rate with control valves FCV1 72% open and FCV2 80% open.
To provide the best control, good operating practice suggests having FCV1 and FCV2 operate between 25% and 75% open. The new maximum rate has FCV1 just within the good practice range and FCV2 just outside that range. However, little control flexibility is available to handle temporary excursions to higher rates.
The second question is pump performance. The hydraulic analysis assumes the pump operates on its performance curve. Pump performance often deviates from that curve. After long in-service time, delivered pump head may differ by up to 10% from that shown on the performance curve.
The system with the base modifications can’t effectively handle higher rate excursions above the average value nor could it run at the desired rate if the pump operates with heads below those documented on the pump curves. Increasing the operating margin requires removing more pressure drop from the system or adding more head.
Reducing pipe pressure drop gives little benefit. The piping system has no specific hydraulic choke points. Achieving pressure drop savings from pipe changes would necessitate extensive pipe modifications for modest benefit.
The two new orifice plates incur a 5.1-psi pressure drop. Using a lower-pressure-drop measuring instrument could reduce this. Ultrasonic flow meters might be a good choice.
Replacing exchangers E1 and E2 could save some pressure drop. However, this would be an expensive change for a modest gain.
Increasing the head available requires pump modification or replacement. Pump P1 has a less-than-maximum-diameter impeller. Switching to a larger diameter impeller would add more than 10% head capability. Unfortunately, the motor on P1 is too small for a larger diameter impeller. Installing a larger impeller also demands a larger motor.
Regardless of the solution, the plant must accept some combination of:
• potentially lower-than-desired flow rates;
• further capital investment to lower pressure drop; or
• additional spending to increase head available.
ANDREW SLOLEY is a Chemical Processing contributing editor. You can email him at ASloley@putman.net