Breakthrough Monitoring or Scheduled Change-Out? Two Ways to Keep a Carbon Adsorber in Compliance

By the NextAir Systems engineering team. Last reviewed October 2026.

A carbon bed does not announce that it is full. One day it removes 99% of the solvent and a few days later it removes much less, and the only way to tell is to look at the outlet or to count the days. Both are legitimate. Which one you pick depends on how predictable your stream is and what a surprise would cost.

This guide sets out what the regulations we have seen actually say, and gives a simple model for comparing the two approaches.

The short version

• One state air rule requires a carbon adsorber to continuously monitor and record the VOC concentration of its exhaust to detect breakthrough.^[1]
• The same rule allows an alternative: replace the carbon at a regular predetermined interval shorter than the replacement interval calculated from maximum design flow and VOC concentration.^[1]
• A schedule is simple but wasteful. Changing a bed at 24 days when it would have lasted 30 wastes about 20% of the carbon.
• Monitoring uses the carbon fully and catches surprises, at the cost of an instrument that has to be kept calibrated.

What the rule says

The Wisconsin Administrative Code, in its section on organic compound emissions, sets out the two routes for a carbon adsorber. The first is to continuously monitor and record the VOC concentration of the exhaust gas stream, to determine whether breakthrough has occurred. The second, offered as an alternative to daily monitoring, is to replace the carbon with fresh carbon at a regular predetermined interval that is shorter than the replacement interval determined by the maximum design flow rate and the VOC concentration in the stream.^[1]

We cite this rule as an example, not as the rule that applies to your plant. Requirements differ by country, state and permit, and yours may add reporting, calibration or alarm duties. The structure is common, though: monitor the outlet, or change the bed before it can fail. The EPA carbon adsorber chapter says the technology achieves 95 to 99% removal when properly operated and maintained, and the monitoring is part of what that phrase means.^[2]

The two approaches side by side

Setting a safe interval

The schedule has to be shorter than the calculated life, and the calculation has to use the maximum design flow and concentration, not the average. Work out the bed life as in our bed life guide, then apply a margin. The right margin reflects how much you trust the inputs. If concentration is steady and humidity is low, a margin of 10 to 15% may be enough. If the line changes products, or the climate is humid and the exhaust varies, the margin should be larger, or you should monitor.

Worked example. A bed with 1,800 kg of carbon is calculated to last 30 days at maximum design conditions. Set the schedule at 24 days, a 20% margin. A year then needs 15.2 changes. Without the margin, at 30 days, it would need 12.2. The margin costs 3.0 extra bed changes a year, or about 5,475 kg of carbon. A monitored bed on the same stream would run until the detector trips, and the saving in carbon has to be set against the cost of the detector and its calibration.

What the margin costs

The 20% margin in the example adds 3.0 bed changes a year, which is 5,475 kg of carbon. Whether that is worth a detector depends on the carbon price, which you should replace with your own quote. For illustration only:

If the annual cost of the margin is larger than the installed and maintained cost of an outlet detector, monitoring pays for itself on carbon alone, before counting the value of catching an excursion early. If the stream is stable and the carbon cheap, the schedule wins. Neither answer is universal, and it is worth doing the sum once.

Monitoring done properly

• Choose the detector for the compounds. A photoionization detector responds differently to different solvents, so calibrate to the compound you have, or to a defined response factor.
• Set the alarm below the limit, not at it. Breakthrough rises quickly once it begins, and you want time to change the bed.
• Keep records. The alarm, the calibration check and the change-out date are the evidence an inspector will ask for.
• Do not forget the inlet. EPA’s adsorber inspection material lists keeping the inlet concentration at or below 25% of the lower explosive limit among the checks.^[3]
• Consider a guard bed. A second bed in series turns breakthrough from an emergency into a scheduled job.

How humid air shortens the interval is covered in the humidity guide, which is often the reason a fixed schedule built on dry-day figures fails in a wet season.

Where this fits with NextAir

Continuous outlet VOC monitoring, with PID or FID detection, is an option on our NVA adsorbers, and so is a multi-stage bed in series. Tell our dry VOC adsorber manufacturer team which route your permit requires and we will build it into the design. Our overview of dry VOC adsorption covers the rest, and the paint and coating page shows a typical stream.

Sources

1. Wisconsin Administrative Code NR 419.045. Control of organic compound emissions: carbon adsorber monitoring.
2. US EPA, Office of Air Quality Planning and Standards. EPA Air Pollution Control Cost Manual, Chapter 1: Carbon Adsorbers (7th edition).
3. EPA / Air Pollution Training Institute (airknowledge.gov). PERM140 course slides: VOC controls.

Figures and tables on this page are calculated or compiled by NextAir Systems and you are welcome to reuse them with a link back to this article.

Frequently Asked Questions