My company wouldn’t buy a quality new air compressor. Instead, it opted to save money by purchasing an almost new unit. However, that compressor in previous service used hard well water. So, we definitely had to descale its tubes.
We started descaling using inhibited hydrochloric acid but were getting nowhere in cleaning the tubes. The water was lukewarm, which didn’t help. So, after a day, I increased the inhibited HCl concentration slightly. The solution from the cleaning tank then became opaque green. My blood pressure exploded! Clearly the procedure was eroding copper out of the tubes of the air compressor intercooler. Nevertheless, I gambled and we continued. I really didn’t know what I was doing but it worked. A flush and a pressure test later and we were ready to go. Sometimes, you’re just lucky.
So, looking back after 25 years, how would I have approached descaling differently? I’d start by scraping samples from the tubes. I would sample the tubes in several locations and separately test each sample with the various descaling agents under consideration.
However, before those tests, I’d check each sample for oils: soak the sample in acid and look for soap slime. Acid plus oil means soap residue — a real mess; aliphatic chains have active groups like aldehydes, ketones, carboxylic acids. If you find oil and organics, as the first step in cleaning, remove them with detergent followed by a hot water rinse.
Then, you must consider solids. HCl, the least expensive pure acid available, is very good at removing PO4-2, CO3-2 and biological slimes. In addition, it is extremely effective at getting rid of iron scale at 140–170°F, i.e., pickle bath temperatures. However, HCl won’t remove sulfates such as CaSO4, silicates, clays, coke, sulfur, pyrite or FeS2.
[callToAction ]
HCl, when properly inhibited, is effective for cleaning carbon steel, cast iron, admiralty brass, bronze, copper-nickel alloys and Monel. It isn’t appropriate for stainless steels, though.
The degree of descaling depends upon inhibitor concentration, average acid concentration, cleaning time, temperature and velocity. Usually, you should restrict velocity to 1–2 ft/sec and limit HCl concentration to a maximum of 10% by weight (5% is recommended). Ideally, keep circulating times to 30 minutes or less at pickle bath temperatures.
Circulation isn’t always needed. Indeed, a “fill and dump” approach may work well if a lot of scale is expected; in such a case, finish with a circulation step followed by neutralization with a hot rinse of 2% soda ash at 140°F for an hour; sparge with nitrogen to drive out the oxygen.
A chelating agent, such as citric acid for food-grade applications or ethylenediaminetetraacetic acid (EDTA) for other operations, could help capture the scale and keep it in solution.
Use of a corrosion inhibitor is key to protecting steel. It’s also a good idea to avoid air. Even with inhibitor, pitting and etching may occur. Inhibitors work through a number of chemical mechanisms such as polymerization, weak Van der Waals bonding and stronger chemisorption bonds — the latter are most effective. Of the various families of inhibitors recommended for HCl cleaning, quinolones generally are the best choice because of the chemisorption bonds formed. In addition to an inhibitor, the cleaning solution usually should include a surfactant or wetting agent, typically at a 0.1–0.2% concentration, to get at grease and organics hidden by the scale.
Don’t automatically opt for HCl; assess alternatives. For instance, sulfamic acid (H3NSO3), which is more expensive, is considered safer; it won’t produce Cl2 if mixed with bleach. However, it is less effective than HCl. A common mixture of 10% H3NSO3/5% NaCl works well for removing Fe2O3. Moreover, H3NSO3 can descale most types of stainless steel, except for some in the 400 series.
Passivation should follow descaling. Copper and aluminum form tight oxide layers when exposed to air; refer to the Pillman-Beckwith correlation for oxide protection.
Stainless steels with more than 12% chromium, e.g., types 304 and 316, self-passivate, but there are still risks at grain boundaries and welds. Passivation is a topic unto itself.
