Fenceline odor monitoring. One node reads the gas, the dust and the wind.
Odor complaints never show up with a timestamp you can trust. They show up as a call to the county, two days late, from a house half a mile downwind. Fenceline monitoring puts continuous readings on your property line, so you know what crossed it and when. SmartCompost nodes read odorant gases and particulate together, with wind measured at the post itself. One record, one clock, ready to hand an inspector.
What fenceline monitoring does for a plant
Stack testing tells you what a control device did on the day of the test. Fenceline monitoring tells you what is leaving the site right now.
It watches the line, not the stack
Nodes sit on the property line and read what crosses it. Every shift, every weekend, 3 a.m. included. Nobody has to stand at the fence with a clipboard and a grab sample.
It runs when nobody is looking
Odor events don't schedule themselves around your work week. A blower trips. A biofilter dries out. A load of bad feedstock hits the tipping floor at 4 p.m. Friday. Continuous readings catch the ones a quarterly survey never sees.
It puts a record behind a nuisance complaint
A complaint is a description: rotten eggs, worst around dinnertime. Your record is a timestamp, a node, a compound and a wind direction. You can act on the second one.
It separates your site from the road and the plant next door
Put a node upwind and you have a background reading. If something blew onto your property before it blew off, the data shows it. That matters most on industrial roads where three operations share one complaint.
It shortens the argument
Most odor disputes run on memory and opinion, and both sides wear out. A continuous record moves the conversation from what people believe to what the instrument logged.
Odorant gases, four particulate fractions and wind on one node
Most perimeter equipment picks a lane. Dust monitors count particles. Gas monitors read two or three compounds. Met data comes from an airport twelve miles away. You end up stitching three datasets and three clocks together, usually the week an agency asks for them.
- Hydrogen sulfide, ammonia, methane, TVOC, CO2, mercaptans, trimethylamine, DMS and DMDS. The compounds people actually smell at a compost pad, a headworks, a rendering cooker or a working face.
- PM1, PM2.5, PM4 and PM10 out of one enclosure. PM10 tracks the visible dust off a tipping floor, a screen deck or a windrow turner. PM1 and PM2.5 are the fractions that come up when the health argument starts at a public meeting.
- Wind speed, wind direction, temperature and humidity read at the post itself. Odor rides the wind at your fence, not a regional average logged at the runway across town.
- Gas, dust and wind share a row. Pull one line of data and you see the whole minute. No aligning exports, no guessing which clock drifted.
- Dust complaints and odor complaints come from the same neighbors, often on the same day. Two systems give you two partial stories. One node gives you the event.
How we pick where the nodes go
Placement decides how much your record is worth. A node in the wrong spot answers a question nobody asked. We walk the site, on paper first and then in person, before we propose a layout.
Start with the wind, not the fence
Pull several years of local wind data and find the prevailing direction by season. Plenty of sites run one pattern in the daytime and another one overnight. The overnight pattern is usually the one generating calls.
Mark the receptors
Houses, the school, the ballfield, the office park, the road where people sit at a light. Draw a line from each one back to the site. Where those lines cross your property line, you want a node.
Mark the sources
Tipping floor, active piles, screening, leachate pond, headworks, DAF, cooker, biofilter discharge, working face. A node downwind and close to the biggest source gives you warning before anything reaches the line.
Put one upwind
The background node is the one people skip and later wish they had. Without it you can show a reading. With it you can show a difference. That gap is your answer when somebody points at your gate.
Look at the ground, not just the map
Terrain, buildings and tree lines steer the plume. Cold air drains downhill at night and carries odor with it. A node on top of a berm can sit above the layer the odor is actually moving in.
Confirm power and signal
Line power is easiest. Solar works where a trench makes no sense. Check cell coverage at the post location, not at the scale house, because the far corner of a landfill is usually a dead spot.
Leave yourself room to move one
The first three months of data usually show one node in the wrong place. Plan for it. Moving a post is cheap. Defending a badly placed record is not.
How many nodes a site usually needs
These ranges cover most of the sites we quote. Acreage, terrain and the distance to the nearest receptor push a site toward the top or the bottom.
- A compact pad with receptors on one side sits near the bottom. A plant with headworks, digesters and dewatering spread down a long parcel sits near the top.
- These sites usually have several distinct sources and a short setback to somebody's property. Cooker area, raw receiving and the DAF often want coverage of their own.
- Long perimeters, a working face that moves, gas wells, and complaints from more than one direction. Node count follows perimeter length and how many sides have receptors on them.
- Receptors on more than one side. A wind pattern that flips between day and night. Terrain that channels air. Multiple permitted sources. A complaint history you already have to answer for.
- One dominant wind direction, one cluster of neighbors, a compact site and no open enforcement file. Start smaller, run a season and add nodes where the data points you.
What the record shows when a complaint lands
The call comes in Monday. The event was Thursday at 6:40 p.m. The inspector wants to know what was running. A logbook and somebody's memory won't settle that. Continuous data will.
Was the wind even pointed at them?
Wind direction at your fence at the time of the complaint, off your own node. If the wind was out of the northeast and the caller lives northeast of you, that is the first thing anybody needs to know.
Did anything actually cross the line?
The reading at the node nearest that receptor, for that window. Not a daily average. The minutes the question is about.
What was running at the time?
Line the reading up against your operating log. Turning, a load received, a scrubber down, a cell left uncovered. That is where the cause usually turns up, and it takes minutes instead of a week.
Did it clear after you fixed it?
Corrective action is easy to claim and hard to prove. A curve that drops after you act, and stays down, is the proof. Agencies notice that. So do neighbors at a public meeting.
Was it your site at all?
The upwind node reads what arrives at your property before anything leaves it. On a shared industrial road that reading can end the discussion in your favor. It will also tell you honestly when the source was yours.
Can you hand it over?
Export the window as a timestamped file, unedited, with the met data attached. No screenshots, no retyping. What the instrument recorded is what the agency reads.
How fenceline data lands in a permit condition
Continuous perimeter monitoring turns up in air permits, consent orders, settlement terms and county siting conditions. The wording changes from one state to the next. The structure repeats.
An action level with an averaging window
The condition names a compound, a threshold and a time window. Your record has to be at that resolution or finer. A fifteen minute action level needs fifteen minute data, not a daily average.
Investigate and correct inside a set time
Root cause and corrective action within a fixed number of hours after an exceedance. That clock starts when the reading happens, not when somebody opens a laptop Monday morning. Alerts have to reach a phone.
Complaint logging and response
Log the call, respond, document what you found, keep the file. Fenceline data drops straight into that file and gives the response section something better than a narrative.
Data retention and access
Keep the raw history. Some conditions ask for a public-facing view. Unedited continuous history is what makes a record hold up in a hearing months later.
Bringing it up before somebody makes you
Plants in siting, permit renewal or expansion often put monitoring on the table themselves. Negotiating a program you designed beats inheriting one written by the attorney across the table.
Where we stop
We are not a certifying body and we do not write your permit. If your condition names a reference method or an approved instrument, confirm it with your permit writer. Our data supports that program. It does not replace anything written into it.
What we are, and what we aren't
You'll ask these on the first call. Here are the answers up front.
Where we come from
SmartCompost has run continuous odor and particulate monitoring since 2019, at compost pads, treatment plants and heavy odor operations.
No US installations yet
We are entering the US market now. If you sign, you are early. We would rather tell you that on the first call than let you find it on a reference check.
No EPA, MCERTS or air district approval
Our equipment does not carry US regulatory approval and we will not suggest otherwise. Treat it as continuous operational and defensive data, alongside whatever your permit requires.
How we read odor from gas data
We convert gas readings into odor relevance with one documented reference method, applied the same way at every site. We'll walk your engineer through the approach on a call.
Which unit goes where
Smart.Odor is the perimeter node described here. Smart.Compost is the fixed multi-gas unit for inside the process area. Smart.Mini is portable, for walking the pad and running spot checks.
Pricing
No price list. Node count, siting, power and connectivity drive it, and those come out of the site walk. Send us a plot plan and your complaint history and we'll size it.
Frequently asked questions
What is fenceline odor monitoring?
Fenceline monitoring measures the air at your property line instead of at a stack or inside the process. Nodes sit on the perimeter and log odorant gases, particulate and wind around the clock. The point is knowing what left the site, at what time, headed which way. That record is what you use for complaint response, for operations and for permit obligations.
How is this different from a dust monitor or a gas detector on a pole?
A dust monitor counts particles and stops there. A gas unit reads a few compounds and stops there. Met data usually comes from an airport miles away on a different clock. Our node reads odorant gases, PM1 through PM10 and local wind at one post, on one timestamp. When somebody calls about smell and dust in the same breath, you have one record covering both.
Do I need my own wind sensor if the airport already reports wind?
Yes. Airport wind is a regional reading taken over open ground, often miles from your fence. Buildings, berms, tree lines and terrain change direction and speed at your property line. At night, cold air drainage moves odor in a direction the regional reading never shows. If wind data is going to carry weight in a complaint file, measure it where the odor is crossing.
How many nodes will my site need?
Composting facilities and treatment plants usually land between 3 and 8. Rendering and RDF operations run 4 to 10. Landfills run 5 to 15. What moves you inside that range is perimeter length, how many sides have receptors, terrain and how many separate sources you are covering. We size it off a plot plan, a wind rose and your complaint history.
Will this satisfy my permit condition?
Ask your permit writer, and ask before you buy anything. If your condition names an approved reference method or a specific instrument, that requirement stands, and our system supports it rather than replaces it. Plenty of conditions describe continuous perimeter data, action levels and documented response without naming hardware. Those are the ones this fits directly. We do not carry EPA, MCERTS or air district approval, and we will put that in writing.
A neighbor two miles away is blaming us. Can the data help?
Usually, yes. You get wind direction at your fence for the exact window they described, plus what your downwind node read and what your upwind node read. If the wind was carrying away from them, that is on the record. If something did leave the site, you find out first and can act before the agency calls. Both beat guessing.
What does getting started look like?
A call, then a site walk. We look at wind history, receptor locations, source locations, terrain, power and cell coverage, then propose a layout with a reason attached to every position. After installation, expect to move one node in the first season once real data comes in. Send us a plot plan and your complaint log and we'll start there.
Size a fenceline network for your site
Send us the site layout, the receptors that call, and what runs overnight. We will propose where the nodes go and why.
Talk to our team →