In the early 1990s, the Jackson Pike Wastewater Treatment Plant in Columbus, Ohio, upgraded its single-stage anaerobic digestion system.

 

That change, aided by the industry’s first large installation of multi-port sliding valve mixers, plus improved heating and continuous sludge feeding, would enable six digesters to do work that required 16 digesters under the earlier two-stage process.

 

But boosting the new system to peak efficiency depended on finding the right recipe to feed it. After six years of experimenting, with the help of automatic sampling, Jackson Pike found the key to fine-tuning its new system, and a new way to cut costs and nearly eliminate the plant’s most time-consuming and least pleasant task: collecting sludge samples for analysis.

 

The automatic samplers gave the plant an accurate way to assess the solids content of primary and waste activated sludges fed to the digesters. That in turn helped the staff determine the optimum mix of sludges for efficient digestion.

 

Finding the blend

The Jackson Pike plant, built in 1937, has an average flow of 80 mgd. Sludges harvested from primary and secondary clarifiers are processed by anaerobic digestion. About 15 percent of the resulting digested biosolids is dewatered by centrifuges to 10 percent solids and hauled away in tank trucks for land application. The remainder is centrifuged to 20 percent solids and incinerated, producing ash that can be recycled into various commercial uses, such as cover material.

 

Bringing the new digesters up to peak performance was complicated because primary and waste activated sludge (WAS) are digested together. “Having committed to single-stage digestion, our big challenge was to maximize WAS digestion using the plant as-is, with no substantial equipment changes,” says assistant plant manager Doug Wise.

 

Primary sludge and WAS initially were piped to a mixing vessel, and whatever blend occurred there was fed to the digesters. “Sometimes it worked well and sometimes it didn’t,” he says. “We didn’t know when it would or wouldn’t, or why.

 

“We get highest efficiency when the solids ratios in both sludges are correctly balanced, so digestion of both components will be completed at the same time. If there’s too much WAS, it gets only partially digested and foams up into the digester’s fabric canopy. We refer to that as the digester getting sick.

 

“Our first step toward learning how this works was to re-route the two sludges into separate holding tanks so we could vary the mixture and study how different solids ratios affected digestion. Then we experimented with many different blends of WAS and primary sludge, also trying them on different types of anaerobic digestion to see what technology would work best with our plant equipment. For that, we needed to draw and test a lot of samples.

 

“The key to pushing the system to its limit without overloading it is knowing the solids ratios of both sludges going in, and controlling the solids ratio of the blend by varying the proportions of both sludges.”

 

Nuisance task

For the first few years, samples were gathered manually: Someone went out every two hours to turn a spigot, fill a cup and dump it into a jar to build up a 24-hour composite sample for analysis. “With about 28 sampling points in our sludge handling stream, this became the most time-consuming aspect of our work,” he observes.

 

“It’s also stinky and sloppy, because when you open a line under pressure, the sludge just shoots out, especially at spigots installed in pump volutes where the pressure is greatest,” Wise notes. “You can’t just crack the spigot open a little, or the sample just gets the thin liquid part; not a true representative grab of the solids. In other words, if you try to minimize the mess, you end up minimizing the integrity of the sample.”

 

Making matters worse, each composite was the work of different people across three shifts, and each person sampled differently. “After several years of experiments based on manual sampling, we saw that reaching an accurate conclusion would require more latitude in blending the digester input, and more reliable sampling of the material going into and coming out of the digesters.”

 

For that purpose, a blending manifold was built to allow three pumps on both the primary sludge and WAS lines to feed a pipeline to the digesters or a pipeline bypassing the digesters. Used simultaneously or in selective combinations, three pumps would allow fine control over the proportions of both sludges going into the blend.

 

The bypass is needed, Wise adds, because sometimes there is too much WAS for the amount of primary sludge available. The undigested leftover WAS then can go directly to incineration, but only if mixed with digested sludge in proportions small enough to prevent the blend’s dewatered cake from being less than 20 percent solids.

 

Automatic sampling

To assess the sludges accurately, the plant staff proposed using automatic ISOLOK samplers built by Sentry Equipment Corp in Oconomowoc, Wis. — two on the lines coming out of the primary sludge and WAS holding tanks leading to the manifold, two on the digester inlet feed and bypass lines leaving the manifold, and one on the line from the digesters to the centrifuges, downstream from where any bypass material would join the digester outflow.

 

“With that setup, no matter what mix we put together or where we send it, we can accurately monitor solids content,” Wise says. The samplers mount onto an access port in the side of a pipeline, tank or pressure vessel. The sampler body encloses a plunger positioned to extend through the port and into the process stream when activated by a pneumatic cylinder (see diagram above).

 

Near the plunger’s tip is an annular ring (the sample spool), sized to capture a measured volume of material by positive displacement — in this application 8 cc per cycle. When the spool darts into the product stream, then withdraws, it brings out a sample of process material and drops it into a closed container beneath the sampler body.

 

Heavy elastomer seals around the plunger at both ends of the spool keep the port closed regardless of plunger position, and the sample remains contained within the sampler body until discharged.

 

The samplers are actuated pneumatically from a control panel. Cycle time, user-programmed within the panel, was set at six minutes to gather enough samples to nearly fill a 2-liter bottle within 24 hours. The panel also directs a pulse of compressed air through a port above each sampler’s discharge opening to eject any sluggish material within the sample.

 

The eject air enters through a three-way valve that doubles as a flushing device. When the valve handle is turned 180 degrees, the airline is blocked and water is allowed to wash through the sampler body to clean out any residual build-up inside.

 

“Automatic sampling done consistently every six minutes obviously produces a much more reliable composite than manual sampling done inconsistently every two hours,” Wise says. “It also reduces labor costs considerably.”

 

Substantial savings

A study conducted by a plant team credited the five automatic samplers in the blending/digesting system, plus two others monitoring the initial sludge outflow from the primary clarifiers, with savings of more than $81,000 per year in labor. Each sampler requires an operator visit only once a day to harvest the composite sample. Lab testing primarily checks total solids and volatile solids.

 

With closely controlled variations made possible by the blending manifold and sampling system, the plant was able to explore alternative technologies of thermophilic digestion and acid fermentation while testing a wide range of blend ratios. In the end, they found that the standard mesophilic digestion used previously — and widely used throughout the industry — would work best with the digesters if controlled properly.

 

“For our system, we now know we must keep the sludge solids content balanced by weight at 30 percent WAS to 70 percent primary sludge,” Wise says. “Any more than 30 percent WAS makes the digesters start to get sick. Any less WAS lets the system fall off peak efficiency. Continual sampling and testing gives us the tools to keep the system tightly controlled.”

 

Confirming loadouts

Automatic sampling also gave the plant an easy way to verify the amount of digested sludge trucked out. “We not only give it away free, but pay to have it hauled,” Wise explains. “Charges are based not on liquid volume but on the pounds of liquid hauled. The most economical concentration was a liquid with 10 percent solids, which is the target concentration used by the contractor for delivering the desired organic application rates.

 

“By installing an ISOLOK sampler on the loadout line, using a flow-proportional control to pull a sample from every 3,000 gallons pumped out, we generate our own solids weight data to compare with the samples received from the contractor.”

 

About the author

Bill Werra is a business unit manager with Sentry Equipment Corp, a manufacturer of sampling equipment based in Oconomowoc, Wis. This is edited from an article that previously appeared in Water & Wastewater Asia magazine.