Temporary Flaring and Output SO2: Alberta AER Permits, Dispersion Modelling and Risk-Based Criteria

Temporary flaring can look simple: burn the gas, keep the operation safe and stop when the job is finished. The air-quality assessment is more complicated.

Output SO2 can mean the mass emission rate leaving the flare, such as grams per second. It can ALSO mean the predicted ground-level concentration at a receptor, usually reported in micrograms per cubic metre. Those are related quantities with different regulatory implications.

For Alberta operators, the practical question is: Can the proposed temporary or nonroutine flaring operation be conducted under the actual gas composition, flow rate, flare configuration and expected operating conditions while meeting Alberta's air-quality requirements?

The current framework is AER Directive 060: Upstream Petroleum Industry Flaring, Incinerating, and Venting, as revised in 2026.

The modelling workflow

Permit pathway → gas composition → sulphur/SO2 emissions → AERflare screening → dispersion modelling → AAAQO/RBC comparison → source optimization or management plan → permit and operating controls

Does the event require an AER permit?

Directive 060 distinguishes planned temporary events from unplanned nonroutine events. Planned temporary activities can include:

• well testing
• well cleanup
• well servicing
• maintenance blowdowns
• underbalanced drilling
• temporary pipeline depressurization

A temporary permit is generally required when sour gas contains more than 5% H2S or the well is classified as critical sour.

A key distinction applies when fuel gas is added.

Two gas compositions can matter

Permit decision: use the source-gas composition.

Dispersion modelling: use the composition after supplemental fuel gas is added.

The gas-well volume allowance also depends on the well classification and whether the well is tied in. Current Directive 060 provides tiered allowances of

• 600 × 10³ m³
• 400 × 10³ m³ and
• 200 × 10³ m³

...for specified gas-well categories.

The event duration matters as well. Ordinary gas and crude-oil wells generally have a 72-hour cumulative limit, with additional provisions for specified coalbed methane wells and circumstances where an extension is justified.

The permit application therefore needs a coherent event description. Flow, duration and total volume should describe the same physical operation.

What produces the output SO2?

Sour gas commonly contains hydrogen sulphide, H2S. During combustion: H2S → SO2

The starting point for the assessment is therefore the actual gas composition and expected flow. Two flares with the same gas flow can produce very different Output SO2 when their H2S concentrations differ. The calculation can also change materially when:

• fuel gas is added
• flow changes
• heating value changes
• the flare operates over a transient flow profile

This is why the gas composition and operating data deserve careful review before the model is run.

AERflare is a modelling tool, not a black box

Directive 060 requires AERflare-incin for temporary and well-test flaring, incineration and enclosed combustion assessments.

SO2 regulations and strategies

The spreadsheet handles flare-specific source calculations and creates model inputs for AERMOD-based dispersion modelling. Screening is the starting point. Refined modelling can follow when screening is insufficient.

The AER's own documentation recognizes that flare assessment can require expertise in chemistry, thermodynamics, atmospheric physics, meteorology, industrial processes and regulatory requirements.

A flare also does not behave like an ordinary stack. AERflare can derive pseudo-source parameters based on:

• buoyancy flux
• momentum flux
• volumetric flow
• heat release
• effective stack height

That matters because flare plume rise can change substantially with gas composition, fuel addition, flow rate, exit velocity, wind speed and atmospheric stability.

Why source characterization matters

A technically correct AERflare run can still be a poor assessment when the inputs do not represent the actual event.

The critical work often occurs before the spreadsheet is opened.

Fuel gas can change the answer

Fuel gas does more than dilute H2S. Adding fuel gas can change:

• H2S concentration
• total flow
• heating value
• heat release
• exit velocity
• plume rise
• predicted ground-level concentration

Directive 060 specifically recognizes fuel addition as a means of increasing heat release and plume rise where source design needs improvement. A modeller therefore needs the actual fuel-gas composition and the intended fuel ratio.

SO2 concentrations are predicted to be lower when fuel gas ratios are higher

This is one area where a sensitivity analysis can be far more useful than a single model run.

Lower flow does not guarantee a proportional reduction

A lower raw-gas flow generally reduces the sulphur emission rate. At the same time, reducing flow can reduce exit velocity, momentum and heat release. That can reduce plume rise and lower the effective release height.

Red hot Flares

The resulting ground-level concentration can therefore respond in a nonlinear way. Directive 060 explicitly warns that reduced flaring rates do not necessarily produce proportional reductions in predicted concentrations.

That is why reducing the flare rate by an arbitrary percentage should never be treated as a substitute for modelling.

Start with screening

The AERflare screening assessment provides the first regulatory test.

The basic question is: Does the proposed configuration meet the applicable air-quality criteria under the screening methodology?

A passing result may allow the assessment to remain relatively straightforward. A failed screening result does not automatically mean the project cannot proceed. The modeller can evaluate:

• fuel addition
• flare flow
• flare diameter
• exit velocity
• stack height
• operating restrictions
• refined meteorology
• refined terrain and receptors

The objective is to identify a technically defensible configuration that also works operationally.

AAAQO and Risk-Based Criteria

The one-hour SO2 AAAQO (Alberta Ambient Air Quality Objective) is 450 µg/m³. For temporary events, Directive 060 or equivalents in British Columbia and other places, provides a risk-based pathway (RBC) for limited predicted concentrations above the AAAQO. The current criteria include:

• 99th percentile: predicted one-hour SO2 concentration must be no greater than the AAAQO
• 99.9th percentile: predicted concentration must be no greater than 900 µg/m³

Percentiles matter

A raw maximum is not necessarily the regulatory statistic.

The assessment has to use the applicable averaging period and percentile treatment for the event.

A risk-based comparison also does not make an actual AAAQO exceedance acceptable. Actual exceedances remain prohibited.

When the model does not meet the AAAQO

This is where the modelling becomes more valuable...

Regulators have advised Calvin that when nonroutine flare modelling does not meet AAAQOs and the assessment proceeds to RBC comparisons, a flare management plan and/or monitoring may be required to operate.

The current Directive 060 framework supports management measures such as:

• flow restrictions
• fuel-gas requirements
• wind-direction restrictions
• wind-speed restrictions
• atmospheric-stability restrictions
• suspension criteria
• ambient monitoring
• meteorological monitoring

The modelling result can therefore become part of the operating procedure.

The model can become an operating rule

A final assessment may specify when the flare can operate, under what meteorological conditions and when the operator must reduce or suspend the activity.

Where a temporary event still requires the risk-based pathway, Directive 060 provides an air-quality management process for addressing those impacts.

Monitoring has to be tied to a response

Monitoring is useful when it has a defined purpose. A workable monitoring strategy needs:

• Location: where the predicted impact occurs
• Trigger: what concentration or meteorological condition initiates action
• Response: what the operator does
• Authority: who makes the operational decision
• Restart: when the flare can resume

A monitor without a defined response adds data. A monitoring program tied to the modelling provides operational control.

Occasionally we see a case where modelling identified a practical monitoring location, meteorological restriction or flare shutdown criterion. But not too often.

A Calvin example: emergency acid-gas flaring

A recent Calvin Consulting assessment for a sour-gas processing plant illustrates the level of detail required. Two emergency cases were modelled:

• an SRU upset case
• a VRU upset case

The assessment used AERflare, AERMOD, five years of meteorological data and elevated terrain. Background SO2 and H2S concentrations were also incorporated. The SRU upset case involved:

• 76.2 m flare height
• 0.305 m flare diameter
• maximum raw-gas flow of 150.1 × 10³ m³/day
• raw-gas H2S concentration of approximately 87.1%
• variable fuel-gas addition
• approximately 20.8 minutes of calculated event duration

The maximum predicted one-hour SO2 concentration was 316 µg/m³ before background and 321 µg/m³ (including 4.7 µg/m³ background), below the 450 µg/m³ AAAQO.

Flares for acid gas

The corresponding H2S result was 3.9 µg/m³ including background, below the 5 µg/m³ AAAQO.

The VRU upset case produced a maximum predicted one-hour SO2 concentration of 255 µg/m³ including background.

These final numbers look simple. Getting to them required representative gas compositions, transient flow treatment, fuel-gas assumptions, source characterization, terrain, meteorology, receptor selection, background concentrations and the correct percentile treatment.

That is the part of the work a spreadsheet output does not show.

Temporary flaring versus nonroutine facility flaring

These categories should be kept separate.

Temporary and well-test events include well testing, well cleanup, well servicing and underbalanced drilling. Directive 060 provides specific temporary-event provisions.

Nonroutine flaring at permanent facilities includes planned events such as turnarounds and equipment depressurization as well as unplanned upset and emergency flaring.

The applicable modelling method, risk-based criteria and management requirements can differ between these cases. Before opening AERflare, establish which regulatory pathway actually applies.

Calvin Consulting

Calvin Consulting Group Ltd. supports Alberta energy facilities with:

• AERflare and AERMOD modelling
• temporary and nonroutine flare assessments
• AAAQO and RBC evaluations
• flare management plans
• monitoring strategies
• regulatory applications
• post-event modelling

The first step is usually a review of the proposed event, gas composition, flow profile, flare configuration and regulatory pathway.

From there, the objective is to build the smallest defensible modelling assessment that answers the actual regulatory question.

Need help with temporary flaring?

Contact Barry J. Lough at Calvin Consulting Group Ltd. with the proposed flare scenario, gas composition and operating information.

A sound flare assessment should give you more than an Output SO2 number.

Flaring info for AER

It should tell you whether the proposed event is supportable, what assumptions control the result and what needs to change before the application reaches the regulator.

Output SO2 is ultimately an engineering, modelling and regulatory question.

For information on underbalanced drilling, visit AER https://www.aer.ca/rules-and-regulations/directives/directive-036.

Calvin Consulting Group Ltd. specializes in Alberta AEPA (Formerly ERSD and AENV) and AER applications. Here's a little more about Calvin's services: 

Air dispersion modelling

Flaring and incineration

Government applications

Benzene emissions

Thank you.

Common flare modelling mistakes

Treating H2S concentration as the SO2 emission rate - H2S is an input. Flow and combustion determine the sulphur emission.

Using the same gas composition for every purpose - Source gas can determine the permit requirement while the fuel-diluted mixture is used for dispersion modelling.

Treating AERflare as a black box - The spreadsheet performs complex calculations but Professional judgement is still needed to decide whether the inputs and results are physically reasonable.

Assuming lower flow means proportionally lower concentration - Flow also affects plume rise and effective release height.

Treating the maximum as the regulatory answer - Percentiles, averaging periods and background concentrations matter.

Solving an exceedance with one arbitrary parameter - Changing stack height, flow or fuel without understanding the underlying physics can simply shift the problem.

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.

Why professional flare modelling pays for itself

The software is free. The consequences of using it poorly are not.

An unsupported assumption can lead to:

• unnecessary flare modifications
• excessive fuel-gas use
• avoidable operating restrictions
• additional modelling
• permit delays
• monitoring requirements
• supplemental information requests

An overly optimistic assessment can create a larger problem when the actual flare event differs from the modelled case.

Calvin's technical staff come from backgrounds in physics, chemistry and meteorology, with extensive air-quality modelling experience using AERMOD, AERflare, CALPUFF and ABflare.

The value is in determining: 

1. what should be modelled,
2. how it should be represented and
3. what the result means for the operation.

That is much more important than simply knowing where to enter numbers in a spreadsheet.