AER Directive 039: Benzene Emissions from Glycol Dehydrators...A Practical Compliance Guide

AER Directive 039 is Alberta's program for managing and reporting benzene emissions from glycol dehydrators and related benzene sources. For facility operators, the important question is practical: How do you determine the applicable benzene limit, calculate emissions, account for controls and document compliance?

The current edition of Directive 039 was released and became effective in 2025. It replaced the 2018 edition.

This guide explains the process from a technical perspective. It focuses on the parts that can make a surprisingly large difference to the final compliance result: site geometry, gas composition, operating conditions, emission calculations, control efficiency and documentation.

The short version

Directive 039 connects a glycol dehydrator's operating conditions to an annual benzene mass limit. The calculation depends on the facility's equipment, gas composition, throughput, glycol circulation, operating status, control equipment and distance to the nearest applicable development. The work must be completed or overseen by a qualified person.

Directive 039 at a glance

The AER also requires supporting documentation to be retained for three years and produced within 10 business days when requested.

Why benzene emissions matter

Benzene is a carcinogenic compound that can occur naturally in oil and gas streams. When gas containing benzene passes through a glycol dehydrator, some of that benzene can transfer into the glycol and ultimately reach the regenerator still-column vent.

The regulatory concern is exposure at public receptors. Directive 039 therefore establishes source emission limits designed to support achievement of Alberta's benzene ambient air quality objectives. The AER states that dispersion modelling was used in developing the reduced emission targets in the directive.

That distinction matters.

A facility calculation may tell you that a dehydrator could release a certain number of tonnes of benzene per year. Atmospheric dispersion modelling addresses a different question: what concentrations could result at specified receptors under defined source, meteorological and terrain conditions?

The two analyses can work together, but they are different technical exercises.

Two different questions

• Emissions estimation: How much benzene leaves the process?
• Dispersion modelling: Where does that benzene go and what concentrations could occur under the modelled conditions?

Confusing the two can lead to unnecessary work in some cases and an inadequate assessment in others.

Step 1: Determine the distance to the nearest close-proximity development

This is one of the first numbers that should be established correctly.

New Glycol Dehydrator

Directive 039 considers a surface development within 750 metres of a dehydrator emission source to be a close-proximity development. The licensee must determine the shortest distance from the dehydrator emission source to the nearest surface development within 1,500 metres. A high level of accuracy is required where a surface development is within that range.

The measurement is taken from the dehydrator emission source, which can change depending on the control configuration.

For a direct still-column vent, the source is the vent. When vapours are routed to a flare, incinerator, burner or condenser, the relevant emission source can instead be the release point from that control technology.

That makes a seemingly simple question considerably more technical: Where exactly is the emission source?

A change in control configuration can change the point from which the distance is measured.

Current emission limits - For a dehydrator without an appropriately designed flare, incinerator or reciprocating engine, the annual limit is:

Where an appropriately designed flare, incinerator or reciprocating engine provides the applicable control, the after-control limit is 1.0 tonne/year for a close-proximity development and 3.0 tonnes/year where the development is beyond 750 m.

There are additional requirements governing whether a flare, incinerator or engine qualifies for those limits. For example, the directive specifies minimum stack heights for qualifying incinerators and reciprocating engines and refers to Directive 060 performance requirements for qualifying flares and incinerators. 

The difference between 750 metres and 751 metres can therefore have a material regulatory consequence. That is one reason the distance should be established carefully rather than estimated from memory or an old site drawing.

Step 2: Determine the actual dehydrator operating conditions

The emission calculation begins with the process. A useful Directive 039 information package needs considerably more information than a facility name and a gas rate. A typical technical data collection exercise can include:

• facility and licence information
• dehydrator type and installation date
• operating status
• operating days
• number of dehydrators on site
• distance to the nearest applicable development
• inlet gas temperature and pressure
• wet and dry gas flow
• gas composition
• benzene concentration
• glycol type
• glycol circulation rate
• pump make and model
• pump speed
• plunger diameter and stroke
• flash-tank configuration
• stripping gas
• emission-control equipment
• control configuration and disposition of recovered vapours

That list comes directly from the kind of working information required to construct a defensible DEOS calculation.

A spreadsheet can make the final result look deceptively simple. The technical work is in determining whether the information going into that spreadsheet actually represents the facility.

A spreadsheet does not create good data

A calculation can be mathematically perfect and still produce the wrong answer when the gas analysis, flow rate, glycol circulation, equipment configuration or control assumptions are wrong.

The quality-control work happens before the final number appears.

This is particularly important for the extended gas analysis.

The benzene concentration is one input to the calculation. It does not directly equal the annual benzene emission rate. Flow, pressure, temperature, gas composition, glycol circulation and other process conditions all affect the result.

Step 3: Get the gas analysis right

The inlet-gas analysis is one of the most important pieces of the calculation.

Testing for Benzene

The current Directive 039 requires an annual representative dehydrator inlet extended gas analysis within the two months before the DEOS revision date. An alternative schedule may be possible where the gas source and benzene content remain sufficiently stable, subject to AER expectations. The analysis must meet the directive's specified accuracy.

A useful analysis can contain dozens of hydrocarbon and non-hydrocarbon components.

• Those components have to be mapped consistently into the calculation method.
• Totals should reconcile.
• Units need to be consistent.
• Benzene must be represented at a resolution appropriate to the regulatory calculation.

That is an area where professional review has practical value.

A Calvin working spreadsheet used for Directive 039 calculations separates the laboratory gas analysis from the GLYCalc input structure and includes component reconciliation before the calculation proceeds. It becomes more important when the gas-analysis quality or component reconciliation materially changes the calculated benzene emission rate.

The key lesson is simple: A small analytical problem can become a large compliance problem when it is multiplied through a process emissions calculation.

Step 4: Calculate the before-control benzene emissions

Directive 039 requires site-specific before-control benzene emissions to be determined using appropriate simulation software, appropriate site-specific testing or both. The AER identifies GRI-GLYCalc as an example of appropriate simulation software for this purpose. Calculations and test-result assessments must be completed or overseen by a qualified person.

For many conventional glycol dehydrators, the calculation begins with normal operating conditions. Consider an anonymized example from a Calvin working file:

The result is the important part. At these representative conditions, the calculated uncontrolled benzene emissions are about 31.63 tonnes/year.

The applicable limit in this example is 1.0 tonne/year. That gap immediately tells the engineer that process optimization or emission control has to do substantial work.

The number also demonstrates why simply knowing the benzene concentration in the gas is insufficient. A gas stream containing 0.046 mole % benzene does not automatically tell you the annual benzene emission rate. The emission calculation depends on how that gas is processed.

Step 5: Consider optimization before adding equipment

Directive 039 recognizes that optimization of dehydrator operation can sometimes reduce benzene emissions before the vapours reach the still-column vent. That can involve changes to process equipment or operating parameters.

Benzene in here? Maybe.

The engineering question becomes: How much benzene can be removed through sensible process optimization and how much additional control is required?

This is where engineering judgement becomes important. Reducing glycol circulation, changing gas throughput or altering another operating parameter may reduce benzene emissions. The change must still leave the facility capable of performing its intended dehydration function.

A calculation that produces a lower emission rate is therefore only useful when the assumed operating condition is realistic. An artificially low throughput or glycol circulation rate may produce an attractive number on paper while describing an operating condition the facility cannot actually maintain.

The 'right' model input is an engineering question

Normal operating conditions should represent the facility's expected operation. The objective is a defensible estimate of emissions under the conditions the DEOS is intended to represent.

A lower number is useful only when the associated operating condition is credible and supportable.

Step 6: Account for control efficiency

Once the before-control emissions are established, the control technology can be evaluated. Directive 039 assigns maximum control efficiencies to several technologies:

The maximum is only available when the equipment is appropriately engineered and operated. The licensee must account for reductions where the actual configuration or operation cannot achieve the maximum.

This distinction is critical. Suppose a dehydrator has 10 tonnes/year of before-control benzene emissions.

A nominal 90% control efficiency produces: 

10 × (1 − 0.90) = 1.0 tonne/year after control

That result may appear straightforward. Real equipment can introduce another variable: uptime.

Directive 039 provides a method in which the recognized technology efficiency is multiplied by the percentage of time during which the control actually removes benzene. For example:

90% technology efficiency × 80% control availability = 72% effective control efficiency

Applied to 1.5 tonnes/year of before-control emissions:

1.5 × (1 − 0.72) = 0.42 tonnes/year

That is the same type of calculation illustrated in Directive 039 itself. A control that exists physically but is unavailable for part of the year cannot automatically receive its maximum theoretical efficiency.

The VRU example - A working example illustrates why control selection matters. The calculated before-control benzene emissions were approximately: 31.63 tonnes/year

The working configuration identified a zero-benzene VRU with recycle and a claimed control efficiency of 100%.

The resulting calculated after-control emissions were: 0.00 tonnes/year

A 100% efficiency claim has a very specific meaning under Directive 039. The VRU must capture the benzene under normal operating conditions and prevent it from being released to the environment or sent to another control as an emission stream. Where the VRU ultimately sends vapours to another control, the applicable efficiency is based on the final disposition.

This is a good example of why control descriptions need to be technically precise. 'VRU installed' does not tell the regulator enough. The important questions are:

• Where do the vapours go?
• What happens to the benzene?
• Does the configuration actually provide the claimed capture?
• What happens when the control is unavailable?

Those questions determine the compliance calculation, but condensers require a different level of evidence.

Condensing systems deserve special attention. The AER states that condenser performance can vary substantially with system configuration and operating conditions. Increased noncondensable gas flow or methane addition, for example, can significantly reduce the ability of the condenser to remove benzene.

For this reason, Directive 039 requires site-specific verification of condenser performance.

Each condenser system's benzene control efficiency must be assessed individually at least every five years unless otherwise specified by the AER. Without site-specific verification, the assigned benzene control efficiency is zero for compliance purposes.

The testing involves a Total Capture Test or another accepted assessment under Appendix 2.

Testing has its own regulatory workflow. Proposed TCT testing must be notified to the AER at least two weeks before the expected field testing date. Testing results submitted without the required notification will not be accepted for compliance purposes.

This is an excellent example of where specialist experience has direct value.

The challenge is no longer simply running a process model. The engineer has to understand the physical control system, define the test correctly, measure the relevant streams and demonstrate that the results are representative of normal operating conditions.

Step 7: Check cumulative site emissions

Another common trap is concentrating entirely on the dehydrator. Directive 039 also establishes cumulative site emission limits for benzene release sources.

The calculation can include dehydrator emissions and other sources such as tanks. Appropriately designed flare, incinerator and reciprocating-engine exhaust emissions can be excluded under the conditions specified in the directive.

The cumulative site limit is also based on the distance to the nearest close-proximity development, measured from the site boundary. That creates two related calculations:

• Dehydrator limit: based on the dehydrator emission source
• Site limit: based on the site boundary and cumulative benzene sources

Where the relevant distances produce different limits, Directive 039 provides a mechanism for applying the least restrictive applicable limit for the relevant sources, subject to the conditions in the directive.

This is another reason a single spreadsheet cell labelled benzene limit can conceal a substantial amount of engineering and regulatory interpretation.

What happens when a facility exceeds the limit?

The consequences are important.

Directive 039 requires the licensee to immediately report identified noncompliance to the AER and take immediate steps to return the operation to compliance. The directive specifically identifies operational modifications, changes to glycol circulation, pump changes and shutdown as examples of corrective action.

A dehydrator cannot continue operating while out of compliance with its Directive 039 emission limit. That makes the accuracy of the original calculation more than an accounting exercise.

A defensible calculation helps establish whether the facility has sufficient operating margin.

The AER also states that industry is expected to maintain an emissions buffer so operational upsets or unscheduled downtime do not cause an exceedance. From an operational perspective, aiming directly at the limit is therefore a poor compliance strategy.

People's health, environmental protection, and smarter choices all depend on benzene emissions

The DEOS: more than a form

The Dehydrator Engineering and Operations Sheet, or DEOS, is the core document connecting the technical calculation to the facility.

The DEOS records the dehydrator's identity, normal operating conditions, control technology and predicted future benzene emission rate.

The current directive requires the DEOS to be reviewed and updated every 12 months and sooner when the dehydrator is relocated, its operating status changes or its normal operating conditions change. Examples include changes to gas composition, throughput or glycol circulation.

The annual extended gas analysis must be completed within the two months before the DEOS revision date and incorporated into the update. The completed DEOS must also be posted at the dehydrator for AER inspection.

The current AER form has deliberately constrained fields and formatting. The directive requires licensees to use the specified form without changing its structure.

That may sound administrative. It is actually useful quality control. A standardized form forces the calculation to document the variables that matter.

Annual reporting

The DEOS describes expected future operation. The annual inventory describes what actually happened during the previous calendar year.

The annual inventory must be submitted to the AER by May 1 each year and must account for the licensee's operating and non-operating dehydrators. Each dehydrator is reported separately. Control outages that increase benzene emissions must be included in the reported annual mass.

The licensee must retain supporting documentation for three years. That documentation can include:

• DEOS forms
• emissions simulations
• extended gas analyses
• operating-condition records
• Total Capture Test reports
• documentation supporting qualified-person responsibilities

The current directive also requires the annual inventory to include methane released in kilograms for each dehydrator that operated during the reporting year.

This creates a useful compliance principle: The number submitted to the AER should be traceable back through the calculation to the gas analysis, operating conditions, controls and underlying records.

Directive 039 and air-dispersion modelling

This is where a common misconception needs to be cleared up. Directive 039 does not turn every dehydrator calculation into an AERMOD project.

Its core emissions calculation is a source-emissions problem. The directive requires the before-control emissions to be established using appropriate simulation software, appropriate site-specific testing or both.

Atmospheric dispersion modelling answers another question. A dispersion model can take the source emission rate and simulate predicted concentrations at specific nearby locations (receptors) under defined meteorological, terrain and source conditions.

That can become valuable when the project involves broader air-quality questions, such as:

• evaluating changes to a facility
• assessing multiple emission sources
• evaluating a proposed new control configuration
• examining receptor concentrations
• comparing modelled concentrations with applicable Alberta objectives
• supporting a regulatory application
• investigating how source changes affect off-site air quality

The AER itself explains that dispersion modelling informed the emission targets established in Directive 039.

The technical lesson is straightforward: Start with the regulatory question. Then choose the smallest defensible analysis that answers it.

A process simulation, field test, spreadsheet calculation and dispersion model each answer different questions.

Using the wrong tool can create unnecessary cost. Using an inadequate tool can create a much more expensive problem later.

Why professional review matters

Directive 039 explicitly places responsibility on a qualified person for the accuracy and validity of the technical work.

That requirement reflects the real complexity of the task. The challenging part is rarely entering numbers into GLYCalc. The difficult parts are deciding:

• which gas analysis represents the facility
• which operating condition is actually normal
• whether the flow rate is appropriate
• whether glycol circulation is realistic
• whether a control efficiency can legitimately be claimed
• whether control uptime changes the efficiency
• where the regulatory emission source actually is
• which receptor or development controls the applicable limit
• whether cumulative site emissions matter
• whether a control outage must be reflected in the annual inventory
• whether the calculation still represents the facility after an operating change

A sophisticated spreadsheet can automate arithmetic. It cannot replace engineering judgement about whether the assumptions describe the physical facility. That is where the cost of professional expertise becomes easier to understand.

A wrong assumption can lead to unnecessary equipment. A missed control limitation can create a compliance problem.

An outdated DEOS can leave the facility working from assumptions that no longer describe normal operation. An incomplete supporting record can make an otherwise sound calculation difficult to defend when documentation is requested.

Good technical work reduces those risks before they become expensive.

A practical Directive 039 workflow

For a facility preparing or updating a D039 assessment, the workflow can be reduced to nine questions:

1. What equipment do we have? Identify every subject dehydrator and document its configuration.

2. Where is the emission source? Determine the actual release point after considering the control configuration.

3. What is the nearest close-proximity development? Establish the distance carefully.

4. What gas are we processing? Use a representative extended analysis with appropriate benzene concentration and component data.

5. How is the dehydrator operating? Document the flow, pressure, temperature, glycol circulation and other relevant parameters.

6. What are the before-control benzene emissions? Use an accepted simulation method, appropriate testing or both.

7. What control efficiency can actually be claimed? Apply the appropriate technology limit and account for configuration and uptime.

8. Are cumulative site emissions also relevant? Check other benzene sources against the applicable site limit.

9. Can the result be defended? Ensure the DEOS, annual inventory and supporting technical records all tell the same story.

That last question is the one that ties the whole process together.

A final technical perspective

Directive 039 looks simple when reduced to a sentence: Keep benzene emissions below the applicable annual limit.

The real engineering chain is much longer:

Facility configuration → emission source → receptor distance → gas analysis → operating conditions → before-control emissions → control technology → control efficiency → control uptime → after-control emissions → cumulative site emissions → DEOS → annual inventory → supporting records

Every arrow is an opportunity for an assumption to enter the calculation. That is why competent Directive 039 work is more than filling out a spreadsheet.

The value of an experienced consultant is the ability to recognize which inputs matter, test whether they are defensible, select an appropriate calculation method, understand what the result actually means and document the reasoning so the final submission can stand behind the number.

For a straightforward facility, that expertise can prevent unnecessary modelling or unnecessary capital spending. 

For a difficult facility, it can prevent the much more expensive cycle of submitting an unsupported calculation, discovering a compliance problem and having to redo the work under time pressure.

How Calvin Consulting can help

Calvin Consulting Group Ltd. works with Alberta energy facilities on atmospheric emissions, regulatory compliance and air-quality assessments.

For Directive 039 projects, technical support can include:

Data review: Review of facility information, operating conditions, gas analysis and control configuration.

Benzene emissions calculations: Preparation or review of before-control and after-control emission estimates using appropriate process-simulation methods.

DEOS and inventory support: Technical review of the information required for the AER forms and supporting documentation.

Control assessment: Evaluation of control configuration, efficiency and operating assumptions.

Air-quality modelling: Where a project requires an assessment beyond the source-emissions calculation, Calvin can determine whether atmospheric dispersion modelling is warranted and develop the appropriate model.

Regulatory support: Preparation of technically defensible material for discussions and submissions involving Alberta regulatory requirements.

The objective is straightforward: use the right technical tool for the actual regulatory question, produce a defensible result and avoid spending money solving the wrong problem.

Need help with a Directive 039 calculation?

Send Calvin Consulting the facility details, current DEOS, gas analysis and control information.

We can review the existing calculation, identify the assumptions that matter and determine what work is actually required.

For a facility where the difference between compliance and noncompliance can depend on a few critical inputs, getting those inputs right is usually far cheaper than finding out later that they were wrong.

Contact Barry J. Lough, P.Phys., P.Met., EP at Calvin Consulting Group Ltd. for help with Directive 039, benzene emissions calculations or related air-quality modelling.

...for dispersion modelling or help with any air quality questions you might have.  Thank you.

Also: The Best Management Practices for Control of Benzene Emissions from Glycol Dehydrators - http://documents.mx/documents/control-of-benzene-emissions-from-glycol-dehydrators.html 

The Canadian Association of Petroleum Producers (CAPP) explains how technical decisions about benzene-containing processes should be made.  PPE (personal protective equipment) should be used correctly by employees, and monitoring equipment should be used to keep them aware of their surroundings.

Common Directive 039 mistakes

Treating benzene concentration as the emission rate - A gas analysis provides an input to the emissions calculation. The annual mass of benzene released depends on the complete process configuration and operating conditions.

Using an old distance - A new residence or development can change the applicable requirements.

Treating a control's maximum efficiency as automatic - The assigned maximum is conditional on appropriate engineering, operation and configuration.

Ignoring control downtime - Directive 039 requires the increased emissions associated with control outages to be reflected in annual reporting.

Assuming a VRU automatically qualifies for 100% - The disposition of the recovered vapours matters.

Treating condenser performance like a standard software calculation - D039 requires site-specific verification of condenser performance because actual efficiency can vary with operating conditions.

Updating the DEOS without checking the gas analysis - The extended analysis is part of the technical basis of the calculation.

Treating the DEOS as paperwork - The DEOS is the documented representation of the facility's normal operating condition and predicted emissions for the coming period.

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.