Who blocked my pipeline?

Awhile back, I got an unusual phone call from an irrigator. “Rob,” he said, “I think there’s a trolley in my pipeline. Can you locate it?”

When he said “trolley,” he wasn’t referring to the old-style San Francisco cable cars; he was talking about wheeled carts that were used to carry equipment for reinstating cement lining during mild steel cement-lined pipeline construction in the old days. It may sound extreme, but it’s not the first time I had heard of a trolley being left in a large pipeline. Legend has it that it happened during the rehabilitation of a large government irrigation area in South Australia in 1980s. However, it may just have been a believable excuse for unaccounted friction losses in a brand-new 3-foot trunk main.

I agreed to help this client out. The case involved an 18-inch diameter PVC rising main delivering recycled water from a sewage treatment works pumping station to a reservoir 6 miles away.

The problem was that, after a year in operation, there was about a 10% reduction in pumping output. The client assured me that he had tested the pumps and they were OK. Therefore, it just had to be a trolley in the pipe.

I approached this problem much as I had during my time in SA Water, where I had experience during the 1980s locating isolated partial blockages in deteriorated 1930s vintage cast iron cement-lined water supply pipelines.

Recording hydraulic gradients

The first step was to record accurate pressure readings along the pipeline at multiple locations (at least 10) that had been surveyed to provide accurate elevation information. The sum of the pressure reading plus the elevation at each point gave the hydraulic head (termed the peizometric height) at each point. Plotting the hydraulic heads with chainage gives a multiple point hydraulic gradient, or HG, much like in the graph shown in figure 1.

Given that the HG was fairly straight, there was clearly no blockage along the way, which would otherwise be evident by a sudden change in slope of the HG at the position of the red arrow.

So, we determined that the head loss must be due to a general friction buildup in the pipeline. To confirm this theory, we decided to “pig” the pipeline. This involves using the pumps to force two foam cylinders along the pipe from the pump end and exiting into the reservoir. The cylinders, or pigs, are about 2 inches larger than the pipe inside diameter, making them a tight interference fit, and 2 feet, 6 inches long.

The instant improvement in the pipeline friction from pigging was nothing short of amazing. The system head loss had been almost totally restored to original performance, resulting in about a 10% flow improvement from the pump station. So, instead of finding a trolley, we determined that a biofilm was responsible for pipe friction buildup.

Pipeline performance can always be viewed from an energy efficiency perspective.

Pipeline energy efficiency

Pipeline performance can always be viewed from an energy efficiency perspective. Figure 2 shows the biofilm affected (red line) and restored (black line) system curves for the client’s pipeline, before and after pigging.

The increase in system head due to biofilm caused the pumps not only to operate at a higher head, but because of the lower pumped flow rate, some of the pumping was forced into peak electricity tariff. The reduced performance pipeline ultimately accounted for about 15% additional pumping energy costs.

Not everyone has an 18-inch pipeline

Of course, not everyone has an 18-inch pipeline in their irrigation system. So how does that relate to the standard irrigator?

A new 18-inch PVC pipe has a Hazen- Williams friction value of about C=150. When reduced to C=135 (10%) through biofilm buildup, the pipe will have the equivalent of a wall roughness of 6 thousanths of an inch. The same roughness in a 3-inch pipe represents a Hazen-Williams C value of 127. That’s a 16% reduction in flow — or a 32% friction loss increase for the same flow!

To maintain the original flow rate in the 3-inch pipe, 32% more pumping energy is required. And that’s just in the first year.

Energy cost with layflat hose

A case in point was observed in an energy efficiency audit conducted in 2014 on a turf farm in New South Wales. A 600-foot-long 3-inch layflat pipe delivering water to a soft hose boom had a head loss of 85 feet compared with the manufacturer’s rating of 46 feet for the same flow — and with no kinks in the hose! That’s a whopping 85% increase in head loss. Not surprising considering that this layflat was transporting algae-contaminated river water and lay in the hot sun all summer, breeding those little critters on the pipe inside wall.

Calculated in terms of energy consumption, the layflat hose was responsible for 46% of total pumping energy costs through a combination of its small diameter plus biofilm buildup.

Here’s the solution

So, what’s the solution? Move to a larger diameter hose. A 3½-inch hose has a new pipe head loss of only 20 feet/600 feet at the same flow, but when that deteriorates due to biofilm, head loss may rise to only about 30 feet/200 feet instead of 85 feet/600 feet, kinks and fittings excluded.

The pump impeller would need to be trimmed or a VFD fitted to potentiate the energy savings. In some cases, the pump may have to be changed out for a lower head pump.

Everyone has a trolley in their pipeline, and it only gets bigger with time. You can’t get rid of it, but you can control its effects, either through energy-efficient pipeline design in the first place or by “pigging” the pipe to get rid of it!

As for the trolley in my client’s pipeline, the legend lives on. He and I still joke about it whenever we can’t explain a pipeline head loss.

Robert Welke, Dip. Mech. Eng., CIAg, CIAL, is the founder and director of Tallemenco Pty Ltd, trading as WATER PUMPING INSTITUTE, a pumping and hydraulics training body.