GCI TECH NOTES ©

Whether sampling from trucks or tanks the process of getting a representative sample of HWF today seems rather obvious. It was not always that way.

In early 1980 the original test tank farm at the Paulding, Ohio General Portland Cement plant consisted of a 12,000 and an 18,000 gallon tank set up inside an old coal bin. There were no sampling ports on the tanks but there was a sampling port on the pump discharges and another one on the bottom of the basket filter used to filter truck transfers. All trucks came in with their own pumps for offloading. Lab testing was done on an after the fact basis at the Systech corporate office in Xenia, Ohio. Early samples were taken by drawing a portion of the sample from the bottom of the filter basket at the beginning, middle and end of the unloading process. When I started I also began pulling a dip sample from the top of the truck for a visual check. You can imagine the scene that I created when I rejected one of the early trucks simply because it did not “look right”. It turned out to be a truck full of mostly water – just a thin layer of solvent on top. About six years later I did the same thing with the first truck delivered to the HWF operation run by McKesson Envirosystems (which became part of Safety-Kleen a couple years later) at San Juan Cement in Puerto Rico. It was one thing for the facility manager in Paulding to reject a truck based only on a look – it was altogether different for the corporate marketing manager to do the same thing on a truck that was accompanied by a signed analysis from McKesson. Again the truck was mostly water. While it was sometimes claimed that I could tell a truck was bad by just looking at it – that was not the case and the issue of getting a representative sample that would allow us to predict in advance the contents of a tank became a critical project in Paulding as we geared up for compliance with the coming RCRA regulations.

As the new tank farm was built in Paulding I specified sampling ports spaced along the ladders up the side of the tank so we could sample at different levels to verify complete mixing and proper agitation. Early tests demonstrated that it took about 30 minutes after the last transfer to a tank for the tank to be mixed adequately enough so that the a sample drawn from the bottom port on the tank was representative of the entire contents. Of course we also found out quickly about the critical need to flush the sampling port just prior to drawing the sample so that the sample would not come out disproportionately high in metals and ash.

The much bigger challenge was obtaining a representative sample of the truck for testing in the onsite lab prior to unloading the truck. EPA had specified in the early RCRA regulations the use of the COLIWASSA sampler for liquids with multiple phases. This sampler was designed for drums – not trucks. The early sampling tubes delivered by the engineers at the corporate Systech office were true to the EPA design. They were heavy, did not close well, did not drain well, and were impossible to clean. My staff and I did not have kind words for the engineers or EPA at that stage. The next step was to take the metal rod out of the center of the tube and stretch it down the side of the pipe with complicated hinges and closing mechanisms on both ends – an improvement but still very heavy and a pain to use. Getting a sample out of these tubes and into a sample can without making a mess was challenging – and they were still very heavy, long and cumbersome to handle on top of a tank at a time where there was no such thing as sample bridges or safety cables. (The most common early accident in HWF operations was someone falling off a truck while taking the sample.)

I cannot take credit for the insight and invention that led to the sampling tube that nearly everyone in the industry uses to this day. That credit must go to a young man whom I had hired straight out of high school to be our first process operator. Carl Laney deserves all the credit for that important development. As he struggled yet again with a new tube provided by the corporate engineers that consisted of pipe, two ball valves and a complex rod connecting the valves on both ends he came to me and asked if he couldn’t just use a ball valve on the end of a pipe and push it shut against the bottom of the truck. Craig Cape, my assistant manager, and I looked at each other in wonder at why we had not thought of the idea and gave him an enthusiastic “Yes!” Of course at that time the fastest way to spread the word on anything in the industry was to tell a truck driver. I think within 1-2 months every facility in the country was using Carl’s design as the word spread.

So the next time you or one of your people work with that ball-valve-on-a-pipe sampler think of the fact that you could be using an EPA designed COLIWASSA to sample that truck and thank Carl for making the job a whole lot easier.