How Is Air Pollution Caused? The Sources That Don't Behave Like Smokestacks

When people ask How is air pollution caused?, the obvious answer is usually factories, vehicles, fires, dust and other things releasing contaminants into the atmosphere. But from an air quality modeller's perspective, that is only the beginning of the story.

The real challenge is often not identifying that a pollutant is being emitted. It is figuring out exactly how it leaves the source. A tall vertical stack is relatively straightforward. But what about:

• dust blowing from an unpaved road,
• emissions escaping from dozens of small pieces of equipment,
• a horizontal exhaust,
• a vertical stack fitted with a rain cap,
• a storage pile exposed to changing winds,
• several nearby stacks,
• or a flare whose plume behaviour changes as the weather changes?

These sources can behave very differently from the textbook image of a smokestack.

That is where air dispersion modelling for how is air pollution caused becomes less about simply entering numbers into a program and more about understanding what the source is actually doing.

The Short Answer: How Is Air Pollution Caused?

Air pollution is caused when contaminants are released into the atmosphere from natural or human activities. Common sources include:

• industrial stacks and process equipment;
• vehicles and transportation;
• combustion;
• flares;
• unpaved roads and material handling;
• storage piles;
• agricultural activities;
• construction;
• wildfires and other natural sources.

The invisible trail of pollution

But not all emissions are released the same way. A pollutant emitted from a 60-metre stack with a hot, fast-moving exhaust can travel and disperse very differently from the same pollutant released near ground level from a dusty road.

That difference matters.

For an air quality assessment, the modeller must decide how each source should be represented so the model reflects its location, geometry, release characteristics, operating schedule and interaction with the atmosphere.

Sometimes that decision has more influence on the result than the choice between two sophisticated modelling programs. Rather than simply How much is being emitted? we ask How is it getting into the air?

The Air Pollution Sources That Make Modellers Earn Their Coffee

Most people picture industrial air pollution coming from a chimney. Sometimes it does. But some of the most interesting (and difficult) sources are the ones that do not behave like conventional stacks.

1. Fugitive Emissions: Pollution Without a Chimney

A fugitive emission is generally an emission that does not come from a conventional, well-defined point source such as a vertical stack. Examples include:

• dust from roads;
• material handling;
• storage piles;
• wind erosion;
• leaks from equipment;
• open areas;
• storage tanks;
• multiple small releases across a facility.

These sources can be difficult to model because their emissions may change with:

• wind speed;
• material moisture;vehicle activity;
• operating conditions;
• season;
• control measures;
• temperature; and
• other site-specific factors.

A dusty road, for example, does not release the same amount of particulate matter every hour of the year.

Likewise, a storage pile may remain relatively quiet under calm conditions and become a much more significant source during stronger winds.  That requires a specific way to address how is air pollution caused.

The modelling question - The challenge is to select a source representation that reasonably reflects reality.

Depending on the source, that may mean using:

• a point source;
• an area source;
• a volume source;
• or a combination of sources.

Emission estimates may come from site measurements, engineering calculations, inventories or published emission factors such as those developed by the U.S. Environmental Protection Agency.

But this is an area where professional judgment matters. An emission factor can provide a number. It cannot automatically tell you whether the resulting model setup actually represents what is happening at the site.

2. Ground-Level Sources Can Matter More Than Tall Stacks

One of the counterintuitive things about air dispersion is that a smaller source can sometimes produce a larger local impact than a much larger source.

Why? Because release height matters.

A tall stack may allow emissions to disperse over a large volume of air before reaching ground level. A source released close to the ground has much less opportunity to do that. Examples include:

• haul roads;
• parking areas;
• stockpiles;
• material transfer points;
• loading operations;
• open pits; and 
• other near-ground releases.

For particulate matter, these sources can become especially important. Their emissions may also be highly variable. A haul road might produce very different emissions depending on:

• traffic volume;
• vehicle weight;
• road surface;
• silt loading;
• moisture;
• dust suppressants; and 
• weather.

This means that the modelling exercise is not simply: What is the annual emission rate?

It may instead be: When does the source operate, and what conditions cause it to emit the most?

Those are very different questions.

3. Horizontal Stacks and Rain Caps: When the Plume Doesn't Go Straight Up

A conventional stack releases emissions vertically with some combination of:

• exit velocity;
• momentum; and 
• buoyancy.

But industrial exhaust systems are not always that cooperative. A stack may discharge horizontally. A vertical stack may have a rain cap or other configuration that significantly reduces the initial vertical momentum of the plume.

Particulates in air pollution

That does not necessarily mean the plume simply stays at ground level. If the exhaust is warmer than the surrounding atmosphere, buoyancy can still cause it to rise.

The modelling challenge is to avoid giving the plume vertical momentum that the real source does not have while still allowing the model to account for its buoyancy.

AERMOD and AERSCREEN include source options designed for situations such as horizontal or capped releases. CALPUFF can similarly adjust the vertical momentum contribution using a vertical momentum flux factor.

The principle is simple: Do not model a rain-capped or horizontal exhaust as if it were a high-velocity vertical jet. That sounds obvious, but getting the source parameters right requires understanding both the physical source and the assumptions built into the model. Oversights like this form part of how is air pollution caused.

4. Flares: Sometimes the Weather Changes the Source Itself

Flares introduce another level of complexity. With a conventional stack, the source characteristics may remain relatively constant while the weather changes around the plume.

A flare can be different. Flame behaviour, plume rise and effective release characteristics may change with meteorological conditions.

At Calvin Consulting, we regularly account for this by using meteorologically dependent flare emission parameters calculated using algorithms provided by Alberta AER through the AERflare spreadsheet.

This allows the model input to better reflect the fact that a flare is not necessarily the same source under every wind speed and atmospheric condition. That distinction can be important.

A simplified model might treat the flare as having one fixed set of source parameters. A more representative approach can allow those parameters to change with the meteorological conditions affecting the flare.

This is a good example of why air dispersion modelling is not always just an emissions inventory followed by a software run.

Sometimes the atmosphere affects the source characteristics, while the source simultaneously affects the atmosphere. That interaction is exactly the kind of detail that can matter when assessing short-term concentrations.

5. What About Multiple Stacks? Do Their Plumes Merge?

This is an interesting modelling issue. When several nearby stacks release similar emissions, their plumes may interact as they rise.

My thoughts on merging Stack plumes

Some modelling guidance provides procedures for representing plume merging. See the algorithm above from the Alberta Air Quality Modelling Guideline. The underlying idea is that nearby plumes with similar characteristics may combine and behave differently from completely independent plumes.

However, there is also a practical question: Do you actually need to merge them? At Calvin Consulting, we have not encountered a situation where plume merging was necessary for our modelling work.

Our general approach has been to model the individual sources separately. There are several advantages:

• each physical source remains identifiable;
• the model setup is easier to review;
• the assumptions are more transparent;
• individual source contributions remain easier to understand; and 
• modern computers can easily handle the additional computational workload.

In other words, there is little practical reason to simplify a facility by combining stacks merely to reduce computing time.

That does not mean plume merging is never appropriate. It means that, in our experience, modelling the individual sources has provided the most transparent and representative approach for the projects we have encountered.

That is an important distinction. A modelling guideline may provide an available technique without requiring that every available technique be used.

6. How Do You Decide Which Source Type to Use?

This is where a modelling project benefits from experience.

There is no universal answer such as: Dust is always an area source. or: A small leak is always a volume source.

The correct representation depends on the physical characteristics of the release. A modeller may need to consider:

This is also why copying a source setup from another project can be risky. Two facilities may have equipment with the same name but very different physical characteristics.

A Practical Hierarchy for Modelling Difficult Sources

When we are evaluating an unusual emission source, a useful approach is to work through the problem in roughly this order.

Step 1: Understand the Physical Source - Before opening AERMOD or CALPUFF, ask:

• What is actually being emitted?
• Where does it leave the facility?
• In which direction?
• At what temperature?
• At what velocity?
• Is the source continuous?
• Does it change with operating conditions or weather?

Some air models account for every dust source.

Step 2: Determine How the Emission Rate Varies - Some sources can reasonably be represented by a constant emission rate. Others cannot.

Fugitive dust, flares, intermittent equipment and seasonal sources may require a more detailed approach.

Step 3: Select the Appropriate Model Representation - Choose the source type and parameters that best reflect the physical release.

The goal is not to create the most complicated model possible. It is to create a model that is representative, defensible and appropriate for the regulatory question being asked.

Step 4: Check Whether the Results Make Physical Sense - A model can run successfully and still produce questionable results.

This is where experienced review becomes important. Questions might include:

• Does the predicted maximum occur where you would expect?
• Are ground-level sources dominating unexpectedly?
• Does a rain-capped stack behave differently from a similar vertical stack?
• Are variable emissions aligned with the hours when they actually occur?
• Does the model response make physical sense?

The software can calculate concentrations. It cannot always tell you whether the assumptions you gave it were sensible.

How Is Air Pollution Caused? The Modeller's Answer

The simple answer  to how is air pollution caused is: Air pollution is caused when contaminants enter the atmosphere.

The more useful answer is: The way contaminants enter the atmosphere can be just as important as how much is emitted.

A tonne of particulate matter from a tall stack, an unpaved road and a windblown stockpile may all be described as emissions.

But they do not behave the same way. Neither should they automatically be modelled the same way. That is why source characterization is one of the foundations of a useful air quality assessment to determine how is air pollution caused.

When the “Simple Stack” Assumption Doesn't Work

Industrial projects can become more complicated when they involve:

• fugitive dust;
• multiple small emission points;
• ground-level releases;
• horizontal exhausts;
• rain-capped stacks;
• variable operating conditions;intermittent sources;
• meteorologically dependent flare behaviour;or 
• other non-standard releases.

These situations do not necessarily require exotic modelling, but they do require careful thinking about how the source should be represented.

The most complicated-looking model is hopefully not the best model. And sometimes the best solution is surprisingly simple.

At other times, a seemingly simple source turns out to require considerably more work. The key is knowing the difference.

How Calvin Consulting Approaches Complex Air Emission Sources

At Calvin Consulting Group Ltd., our work involves more than selecting a dispersion model and pressing Run.

We start by understanding the facility and the source. For unusual or complex emissions, that means asking questions such as:

• What physically causes the emission?
• How does it vary with operations?
• Does weather affect the release itself?
• Is a conventional stack representation appropriate?
• Are there regulatory methods specifically designed for the source?
• What level of complexity is actually justified?

Our approach is practical. We do not add modelling complexity simply because it is available.

For example, while plume-merging methods are available in some guidance, we have so far found that modelling individual stacks provides a transparent and representative solution without creating a meaningful computational burden.

On the other hand, when the physics of the source genuinely change with meteorological conditions—as can occur with flares—we regularly use the meteorologically dependent parameters generated using Alberta AER's AERflare algorithms.

That is the difference between complexity for its own sake and complexity that actually improves the representation of the source.

Our goal is straightforward - To produce an air quality assessment that is:

• technically sound;
• representative of the actual facility;
• appropriate for the regulatory jurisdiction;
• transparent about its assumptions; and
• understandable to the people who have to make decisions based on it.

If your project includes unusual emission sources, fugitive emissions, flares, capped stacks, horizontal releases or a facility with dozens of individual sources, it is worth discussing the modelling approach before the design is finalized.

Calvin Consulting Group Ltd. provides air quality dispersion modelling using tools including AERMOD, CALPUFF, AERFlare and regulatory modelling methodologies appropriate to the project and jurisdiction.

Don't let air quality worries hold you back. Let's talk! Contact Calvin Consulting today!

​Clean air is our Passion...Regulatory Compliance is our Business.

Why does the air feels thick sometimes?

Or is it just your imagination? I doubt it; the sneaky culprit of air pollution lurks everywhere. We could benefit from exploring how our daily living and the world around us contribute to this threat, though we may not see it right away.

Do you have concerns about air pollution in your area??

Perhaps modelling air pollution will provide the answers to your question.

That is what I do on a full-time basis.  Find out if it is necessary for your project.

Have your Say...

on the StuffintheAir         facebook page

Other topics listed in these guides:

The Stuff-in-the-Air Site Map

And, 

See the newsletter chronicle. 

Thank you to my research and writing assistants, and the author remains responsible for the content.