A greenhouse carbon filter is a possible treatment for a specified gas or odor problem. Before selecting one, identify the contaminant, where it enters or is generated, which air stream needs treatment, and how removal will be checked. Room volume alone does not establish a filter's gas-removal capacity.
This guide covers the selection and verification questions for that decision. It does not establish crop-yield improvement, disease prevention or worker-exposure compliance for a particular installation.
First separate the problem from the equipment
Write a short problem statement: what was observed, where, when, and during which activity. An odor complaint at an exhaust outlet and a suspected contaminant entering through an intake are different tasks. A photograph of a filter or a general room-air reading does not identify the substance responsible.
Record whether the event follows spraying, cleaning, nearby emissions or a change in production. Where exposure or crop injury is suspected, obtain appropriate identification and assessment instead of treating the smell as a concentration measurement. Do not conduct informal smell tests on potentially hazardous air.
The EPA guide to air cleaners distinguishes mechanical particle filtration from gas adsorption. Activated carbon can treat selected gases; it does not remove every gas. The guide also puts source control and suitable ventilation ahead of reliance on an air cleaner. Its scope is residential air cleaning, so it supplies these mechanism distinctions, not greenhouse performance guarantees.
Define the air path before requesting a quotation
Draw the intake, occupied crop zone, recirculation return, treatment unit and discharge. Mark which dampers and fans operate in each mode. Answer three questions before comparing filters:
- Is the treatment on incoming air, recirculating air, exhaust air, or more than one separate path?
- What portion of the relevant flow actually passes through the treatment housing, and where could it bypass the media?
- What other duty must that fan perform, such as heat removal or fresh-air supply?
For recirculation, show where treated air returns and which crop zones the return serves. For exhaust treatment, identify the discharge location and the complaint or emission being addressed. Do not assume that treating an exhaust stream also cleans every part of the crop zone.
Keep ventilation, humidity control and gas treatment as separate duties in the greenhouse climate design brief. A filter is not a substitute for a moisture-load calculation or a pest-management program.
Ask for evidence for the actual contaminant
The California Air Resources Board's gas and odor guidance explains that ordinary particle filters do not remove gases and that specialized gas-removal products must be evaluated for the chemicals of concern. Its residential guidance should not be converted into a greenhouse sizing table. In a greenhouse quotation, request the following application-specific evidence rather than accepting a broad “VOC removal” label:
| Item | What the quotation should identify | What a missing answer prevents |
|---|---|---|
| Contaminant | Named gas or defined mixture and inlet concentration basis | Matching a test result to the actual problem |
| Operating conditions | Temperature, humidity, flow and test duration | Comparing the test with greenhouse operation |
| Treatment assembly | Media type, quantity, bed arrangement and housing | Knowing which configuration was tested |
| Removal record | Inlet and outlet measurements, units and method | Checking whether the claimed reduction was measured |
| End point | Outlet limit or breakthrough criterion and replacement basis | Planning a service interval from evidence |
| Air-side duty | Pressure loss at the specified flow and allowed end condition | Selecting the fan operating point |
Do not substitute a water-treatment iodine number, feedstock description, or room-size label for this evidence. If no matching gas test is available, record the selection as unverified and define a limited evaluation before relying on it for the full site.
Match the fan to the installed resistance
Request the proposed filter pressure-loss data and fan curve together with the actual duct layout. Include prefilters, bends, dampers and any other components in that path. Use the intended installed flow and operating mode in the check; a free-air fan rating does not describe the complete assembly.
At commissioning, record measured flow, pressure across the treatment assembly, fan setting and damper positions. Save that baseline so a later pressure change can be investigated against a known configuration. A pressure check assesses the air path; it does not measure how much gas-removal capacity remains in the media.
If the flow is wrong, investigate the installed path and measurement before increasing speed. A speed adjustment must remain within the supplier's supported treatment conditions and equipment limits. Do not apply a universal percentage allowance for duct losses or an arbitrary one-to-three-minute room turnover rule.
Keep CO₂ enrichment and exhaust losses distinct
Label every measured CO₂ location and every air stream crossing the greenhouse boundary. Exhaust removes air from the enclosure; an internal recirculation loop returns air to it. These are different terms in the enclosure balance. A drop in crop-zone CO₂ therefore does not by itself prove that the carbon adsorbed it.
For an enriched house, record dosing times, vent positions, fan modes and sensor readings on the same timeline. Ask whether the chosen media has relevant CO₂ interaction data if that interaction is part of the proposed design. Do not reduce required ventilation or disable a safety function merely to retain dosed CO₂ or extend filter life. Climate and safety requirements must remain explicit constraints on the operating sequence.
Verify performance before setting a maintenance interval
Agree the acceptance record before installation: the target substance, sampling locations, instrument or laboratory method, operating conditions, measurement duration and pass criterion. Compare inlet and outlet under the agreed conditions. Where the original issue is at an exhaust outlet, include that location in the verification plan rather than testing only a convenient return duct.
Keep air-side checks and contaminant measurements in separate columns. A clean prefilter or unchanged pressure does not establish continued gas removal. Conversely, a flow change requires an air-path investigation rather than an automatic conclusion that the adsorbent is exhausted.
Use the supported replacement criterion and site observations to schedule service. There is no defensible universal 12–24 month interval for every gas, media bed and greenhouse. Record installation date, exposure events, measured performance and changes in operating conditions. Handle spent media according to the adsorbed substance, supplier instructions and applicable waste requirements; do not assume it is suitable for compost.
FAQ
Frequently asked questions
Does activated carbon replace a particle filter or pest screen?
No. Gas adsorption, particle filtration and physical pest exclusion have different functions. Assess each component against its own target. The presence of carbon does not prove removal of spores, insects or crop disease.
Can I size the system from greenhouse volume?
Volume helps describe the enclosure, but it does not establish gas load, media capacity or required contact conditions. Define the target contaminant and air path, then obtain matching treatment and pressure-loss data.
Does a carbon filter guarantee safer air or higher yield?
No. A measured reduction of a named contaminant is a specific treatment result. It is not automatically an exposure-compliance finding or a crop-yield result. Those outcomes require their own assessment and evidence.
When should the carbon be replaced?
Use the supported end-of-service criterion for the target contaminant and operating conditions, alongside measured performance. Pressure, odor and calendar age alone are not a universal measure of remaining adsorption capacity.
