Increases in plant capacity often call for fitting more flow volume through existing lines. A conventional desk-based approach to estimating hydraulic capacity starts with the line geometry, calculates an effective line length, and then uses estimated physical properties to calculate a line pressure drop for the new flow rate.
Too often this approach fails because of significant differences between calculations and reality. Actual plant hydraulic performance usually doesn’t match calculated expectations because at least one of three factors — supply pressure, pressure drop and destination pressure — isn’t what you think it should be. Here, we will focus on supply pressure and pressure drop.
Multiple standards include different acceptance tests for centrifugal pumps. However, testing pumps costs money. Did your pumps get tested before delivery? Next, what are the acceptance levels? The most commonly used standards, those of the Hydraulic Institute and the American Petroleum Institute, offer a choice of acceptance classes. Most often, the values used are ±8% on flow and ±5% on head for most chemical industry pumps and ±5% on flow ±3% on head for oil and gas industry pumps. Other limits apply to power and efficiency.
Wear on pumps always decreases pump head available. Based on my experience, in service you should expect up to 10% less pump total dynamic head than that shown on the pump performance curve. (Achieving more than that requires a higher level of performance and maintenance than the typical plant can consistently provide.) So, the supply pressure may be far less than what you anticipate. Lower supply pressure results in lower flow rates.
[pullquote]
Pressure drop depends upon the geometry of the flow channel, which includes factors such as size, surface roughness, shape, obstructions and length. Tables and graphs of friction factors and equivalent lengths are available for all these but sometimes the line isn’t what you assume. Situations I and others have encountered include pipe schedule differences with segments of thicker pipe in lines, items dropped into pipes and fittings, welding blanket fragments in pipes, plastic trapped in pipes, damaged and pinned check valves, pipe blocked by deposits (see: “Scale Back on Heat Tracing”), temperature solidification deposits, sedimentation deposits, and other factors.
In these cases, the tables and figures of friction factors and equivalent length have no value. The real hydraulic performance differs so much that the calculated possible flow rate never can be achieved.
The solution is simple for many applications. Straightforward checks and instruments can give you the information you need to make a valid estimate of possible capacity.
First, walk down the line. Look for partially closed isolation valves or open bypasses around control valve stations. Operators often use isolation valves to throttle flow to prevent operating problems. If an isolation valve is partially closed, find out why. You may have bigger challenges in increasing plant capacity than just a hydraulic limit. Open control valves or other bypasses signal you already have flow hydraulic problems. Understand why every bypass is being used and what the implications are.
Second, check control valve positions. A valve that’s fully open or close to it shows a flow line is already at or near its capacity. Work on understanding why the valve is so open and how to deal with that.
After those two steps, get an accurate pressure gauge. Run a pressure survey down the line. Use the demonstrated system pressures and flow rates to calibrate plant performance. A pressure reading coupled with the simple relationship that pressure drop goes up with the square of flow rate in turbulent flow will give a better answer for system capacity than the most-complicated flow calculations.
Of course, complex flow patterns such as phase change, partially full pipes, multiple-phase flow, compressible flow, laminar/turbulent transition and non-Newtonian fluids don’t have such simple flow relationships. Physical property changes also can affect this relationship. One example is temperature loss and viscosity changes in long lines.
Nevertheless, the pressure gauge, when correctly used (as explained in “Do Simple Things Right!”), is a tremendous tool for understanding the flow capacity of systems. I suggest you intensely question any proposed plant capacity increase that doesn’t come along with a hydraulic analysis backed up with pressure surveys. Too many plant expansions have failed due to a disconnect between actual conditions and desk-bound calculation.
