Minding the Stack

FROM THE SEPTEMBER ISSUE: Maintaining compliance while optimizing costs and operations is a balancing act for ethanol producers.
By Susanne Retka Schill | August 15, 2019

Mastering the balance between emissions control and cost control is important for an ethanol plant. The targets for emissions control are set in the facility’s air permit, while the costs involve energy, maintenance and compliance testing.

The energy cost in oxidizing volatile organic compounds (VOCs) at 1,600 degrees Fahrenheit can be significant. “The typical exhaust matrix from a dryer at an ethanol plant is 45 to 50 percent water, and we’re talking 30,000 to 80,000 cubic feet per minute (CFM),” says Jordan Laster, vice president of analytical services at Alliance Source Testing. “It takes a massive amount of energy to keep temperatures up to maintain gas phase.”

A well-run plant that is not trying to cut costs on control devices will generally meet emissions requirements without a problem, he says. Turning down the temperature or recycling more process liquids, however, can save as much as $250,000 per year in utility costs. But as the temperature goes down or recycling goes up, the destruction efficiency goes down. “A lot of facilities try to walk the line between the lowest temperature they can run and still meet emissions limits,” Laster says. Some target operating at 50 percent of limits and some shoot for 90 percent of limits. “And they do see a rather substantial difference in their environmental control cost.”

Plants balancing control and cost often turn to stack testers using FTIR (Fournier-transform infrared spectroscopy—Method 320 in the U.S. EPA testing protocols). “It’s ideal for live results,” Laster says. “It’s great for engineering and tuning—what happens when I turn this valve or change this flow rate? Plants can see how that changes emissions on a live basis.”

FTIRs are also used for compliance tests, although Laster says he sees some misunderstanding of the test’s best application. “FTIR is very appropriate for scrubbers and process vents, because these typically have very low volumetric flow rates,” he explains. “If I get a nondetect there, I have to report that detection limit, but I’m multiplying it by a very low flow rate. The resulting mass emission rate ends up being relatively insignificant when compared to the permitted emission limit.”

The most common VOCs detected are ethanol, acetaldehyde, formaldehyde, ethyl acetate and, occasionally, methanol, Laster says. Air quality permits specify an allowable emission rate, generally stated in pounds per hour. The challenge for emission control and testing, Laster adds, is that “what we are looking for is in the order of a couple parts per million and the emissions rates end up calculating to typically one to five pounds per hour of each of these things.”

The better choice for testing high volumetric sources—regenerative thermal oxidizer (RTO) stacks and distillers dried grains with solubles bag houses—is to use a manual method, such as EPA’s Method 18 or a National Council for Air and Stream Improvement method, where the gas is bubbled through reagents and/or adsorbent tubes that are sent off for laboratory analysis. While more costly, primarily because of the required off-site lab work, the greater accuracy with lower limits makes a big difference, Laster says.

The reason lies in how the regulations say nondetectable compounds are to be handled, he explains. “In stack testing, there’s no such thing as zero.” When a testing method returns a nondetect for a certain compound, the number recorded is just below the threshold that the particular test could detect. For example, if a test can’t detect 1 part per million (ppm) or less, a nondetect test result enters a one and not zero. A more accurate test method, with a limit of 0.5 ppm, would record 0.5 ppm. “The worst-case scenario is after a stack test is completed and a bunch of nondetect results are added up, the resulting emission rate is above the permitted limit,” Laster says. “And that can happen, and does happen without proper pretest planning.”

Another consideration when evaluating test methods is the quality assurance/quality control (QA/QC) performance. Laster says he’s found the NCASI self-validating methodology to be much cleaner than EPA Method 18. “It’s very difficult to get all of the QA/QC parameters,” he says. “Even on a very good test, you might have some compounds that don’t meet all QA/QC criteria.” While regulatory agencies typically accept some method quality issues, too many test issues can trigger additional follow up by the agency and potential invalidation of the test report.

Iowa Department of Natural Resources is an example of an agency that tightened its air quality oversight in the past three or four years, Laster says. “They didn’t go back in time, and it didn’t cause any liability to the facilities, but IDNR came down and said, ‘These are the only things we’ll accept, and all tests must be done this way moving forward.’”

Maintenance Considerations
The balance between cost controls and emission targets is only part of the battle. Maintenance and operations also require attention. With many plants running hard, at 110 or 120 percent of capacity, John McDowell, sales engineer with Eisenmann Corp., cautions plants to keep a close eye on their abatement equipment. 

Distillers grains dryer exhaust is a challenging application, he says, with water- and corn oil-laden exhaust coming from the dryers that can condense on surfaces, and high particulate matter (PM) and carbon dioxide levels that can create acids. The maintenance issues differ depending on the type of system used: thermal oxidizers paired with heat recovery steam generators (TO/HRSG) or an RTO.

In RTOs, a ceramic bed acts as a heat exchanger. Exhaust enters the RTO and passes across hot ceramic media and into the combustion chamber operating around 1,600 degrees to destroy PM and VOCs. The exhaust air then flows across a second ceramic bed that captures the heat until the exiting temperature is close to the inlet temperature. Periodically, the flow switches direction so the hot outlet bed becomes the preheating inlet bed and the warm inlet bed becomes the heat recovering outlet bed.  “Over time, the PM in the dryer exhaust air can build up on the ceramic media, creating channels and restricting flow,” McDowell explains. “You may need to replace ceramic media.”

In a TO/HRSG, after passing through the combustion chamber, the exhaust moves through the HRSG’s fin tube bundles to generate the plant’s steam. “Eventually, you will begin to see wear on the fins, with them getting shorter and reducing heat transfer,” McDowell says. “Also, PM starts building up and, unless you remove it, you’ll see a reduction in efficiency and you may need to replace tube bundles.”

While the maintenance issues differ, the primary means of monitoring performance are the same for both systems. One basic measure is to track temperatures over time, observing whether the difference between inlet temperature and stack outlet temperature begins to widen—an indicator that efficiency is beginning to drop. “It’s very measurable in the field,” McDowell says. “Engineering efficiencies can get complicated with mass flow, temperatures and fancy calculations, but any operator can look at this and say, ‘What are my temperatures across the system?’”

Pressure profiles are another indicator, using trends in the force pulling from the dryer upstream and the inlet pressure versus the outlet pressure. “You’ll see an increase in pressure drop across the ceramic bed or HRSG tube bundle as it fouls,” McDowell says. “For people who don’t have pressure gauges in the right places, you can look at fan speed or amps drawn by the big blower. As that goes up, it’s telling you the system is getting fouled.”

McDowell also recommends plant maintenance teams take photos during semiannual cleaning and inspections. Surface degradation and wear become apparent when compared to previous photos. “It’s easy when doing the inspections to overlook something, but when you’ve got the photos, you see what new looks like, compared to today.” 

With plants running hard and taxing their oxidizers, plus an increasing interest in reducing carbon intensity, some are looking at replacing TO/HRSG systems with RTOs, McDowell reports. Separating the plant’s steam generation from the thermal oxidizer can make it easier to optimize both operations. While plants do run well with a TO/HRSG, some struggle. “Not all plants operate at steady state,” McDowell says. “If you’re drying more DDGS, you may overproduce steam. Conversely, if you’re making wet cake and not drying, you may need to over-fire the TO to generate enough steam for the front end and distillation.”

Author: Susanne Retka Schill
Freelance Journalist