Proof-of-Concept Testing: Software To Quantify Methane Emission Rates in Real-Time

SECTION 1 - INTRODUCTION

This Final Project Report represents the culmination of an 18-month methods-development program, sponsored by the Province of Alberta and administered by Emissions Reduction Alberta (ERA) as part of their “Methane Challenge Initiative.” The title of our project was, “Proof-of-Concept Testing: Software to Quantify Methane Emission Rates in Real-Time.” The end-product was a fully integrated, methane emission-rate measurement system (i.e., the “System”), which calculates, in real- time, methane emission rates from certain oil-and-gas (O&G) industry sources.

Drawing heavily from a total of five Major Deliverables (MD’s) spanning this period, this Final Project Report provides an overview and general chronology of the technical tasks leading to development of a Set of Specifications for eventual System commercialization. The software tested is known as e-Calc 2 (emissions calculation, second-generation). Nine days of successful field testing, carried out during August 2018, involved the continual, outdoor release of carefully controlled amounts of methane by our project team member, InnoTech Alberta, at their research facility in Vegreville. Other members of the project team were: Boreal Laser, Inc. (Edmonton), responsible for all methane measurements using their tunable diode laser (TDL) system; Met One Instruments, Inc. (Happaugue, New York), responsible for the design and assembly of the specialized meteorological measurement system used in the field; and Loover Partnership (Morristown, New Jersey), responsible for necessary e-Calc software modification and statistical consulting.

1.1   Document Organization

Table 1-1 identifies the project milestones set forth in our Scope of Work. Also shown are those milestones which required preparation of comprehensive MD’s. Preparation of an ERA Progress Report was also required upon completion of each project milestone.

TABLE 1-1. PROJECT MILESTONES AND ASSOCIATED M AJOR DELIVERABLES

Table 1-2 identifies, for each section of this Final Project Report, the corresponding Major Deliverable which serves as the primary basis of information. All MD’s are included in their entirety as attachments to this report. While the report sections listed in Table 1-2 are generally consistent with the project milestone chronology, Sections 2 and 3 represent necessary departures, as the successful creation of e-Calc 2 was prerequisite to the preparation of the corresponding MD’s.

TABLE 1-2. FINAL REPORT SECTIONS AND CORRESPONDING M AJOR DELIVERABLES

There is another issue concerning Table 1-2 for which some explanation would be helpful. Our original Scope of Work did not envision the need for Section 7, “Supplemental Booster-Station Analysis” (and, accordingly, two iterations of the System specification). The impetus for this extra work occurred during preparation of the Major Deliverable for Milestone F (Section 5 of this report) when, upon further examination of the field data, we believed we had identified a significant issue related to the treatment of the background methane and the meteorological data used as input to e- Calc 2. If this analysis were explored, we knew the results would likely lead to a refinement of the System specification. Therefore, in the Milestone F Report, we committed to perform this additional analysis as part of the Major Deliverable for Milestone H: Set of Specifications (Appendix E).

The additional Milestone H analysis proved successful, providing tangible improvement to the System specification. However, while performing this analysis, another pathway for further exploration presented itself. This time, though, successful results would enable extension of the System specification to the booster station – previously eliminated as a source due to issues with methane background concentrations. This supplemental booster-station analysis was also successful, and is presented as its own section (Section 7) in this Final Project Report. The final System specification is, accordingly, presented as Section 8.

Finally, most sections can be broadly thought of as summaries of the corresponding Major Deliverables. Therefore, the reader is advised to consult the actual MD’s (attachments) for those details not discussed in the main body of this Final Project Report.

1.2   Project Overview

The software tested is generically referred to as e-Calc (emissions calculation). The first-generation version of this software (e-Calc 1) calculates mass-per-time emission rates during daytime hours from ground-level sources. E-Calc 1 employs what is often referred to as “inverse modeling,” based on AERMOD (American Meteorological Society / EPA Regulatory Model) – a U.S. Environmental Protection Agency (U.S. EPA) “Guideline” air dispersion model for regulatory application. Instead of predicting a downwind concentration at a point in space from a known source emission rate (as AERMOD typically does), e-Calc 1 predicts that emission rate from a measured downwind (crosswind), path-integrated concentration and contemporaneous onsite meteorology.

E-Calc 1 can derive emission rates of methane (or any other measured compound) from most ground-based sources. Importantly, this software offers the capability of generating such emission rates in real-time. However, a significant up-front effort is required, prior to field deployment, to enable AERMOD (and thus e-Calc 1) to simulate the vertical wind-speed profile and atmospheric turbulence – critical model input parameters. AERMOD employs what is known as the flux-gradient approach for simulating these input parameters.

Two distinct goals comprised our ERA project. The primary goal was, first, to modify the software to accommodate a more sophisticated and robust treatment of meteorology (i.e., to create e-Calc 2 based on a new version of AERMOD – modified to employ the eddy-correlation approach for simulating the above model input parameters) and, second, to field-test this second-generation version of the e-Calc software, based on carefully controlled methane releases from simulated, leaking upstream sources. The intent was to eliminate the need for the arduous pre-field tasks and make possible the software use during the nighttime.

The four simulated sources were:

1. a booster station, comprised of a compressor engine and a condensate tank;
2. a gas-gathering pipeline assembly;
3. a gas-transmission line; and
4. a production pad.

Only one simulated source was tested on any given measurement day. As mentioned, all controlled methane releases were conducted by InnoTech Alberta, with all field work performed at InnoTech Alberta’s Vegreville R&D facility. Path-integrated methane measurements were performed by Boreal Laser using one of their GasFinder TDL spectrometers; all TDL measurements were made at a height of 1.0 meters above the ground. All meteorological measurements were made using a sonic anemometry system designed and assembled by Met One Instruments. The methane source was compressed natural gas, with a methane concentration of 76.6 percent (760,000 ppmv).

It should be noted that there has never been a performance evaluation of AERMOD based on this more sophisticated treatment of meteorology, nor has the U.S. EPA yet provided the software coding for this model option. In theory, the AERMOD results should be improved (and, accordingly, the corresponding e-Calc predictions); however, such results could not be guaranteed.

The secondary goal, a benefit to the Alberta Climate Change Office (ACCO), was to apply e-Calc (both versions) to essentially re-create the fugitive methane and carbon dioxide emission rates from the Canadian Natural Resources Limited (CNRL) mine-face and tailings pond operations in Fort McMurray, as reported in CNRL’s two latest (at the time) annual submissions on facility greenhouse gas (GHG) emissions. Our analysis used onsite, 15-minute-averaged path-integrated methane and carbon dioxide data, collected across portions of these sources in 2015 and 2016 by CNRL using a Boreal Laser TDL spectrometer, together with onsite, coincident meteorological data and contemporaneous flux-chamber sampling data. The hope was that e-Calc would be demonstrated a viable and attractive alternative to the techniques currently employed for measuring GHG’s from the oil-sands sources, and that the time and cost for GHG reporting would be greatly reduced.

1.3   Market Need

We know of no other measurement system which can, in real-time, generate accurate estimates of methane emissions from ground-level sources. It is difficult for the Province of Alberta to enforce existing methane reduction mandates without an accurate baseline against which to compare. The rapid and inexpensive means of measuring methane emission rates afforded by the success of this Project is clearly a disruptive technology.

When used in combination with the TDL system, e-Calc offers a common-sense approach for prioritizing repairs in the O&G industry, which can reduce product loss while adding bottom-line profit. By quantifying methane emissions from principal source types within a given industrial sector, the quality of emissions inventories should be vastly improved, thereby facilitating an accurate methane baseline against which future reductions can be reliably assessed.

In addition to leaking upstream process components, target markets in Alberta for this System include: (a) municipal landfills; (c) combined animal feeding operations (CAFO) facilities; and (c) major oil-sands sources, consisting of tailings ponds and mine faces. In fact, the feasibility of employing e-Calc 2 to assess methane emissions from these oil-sands sources was demonstrated during the work for ACCO (secondary project goal), thereby laying the groundwork for a proposed field demonstration at a tailings pond.

SECTION 2 - E-CALC 2 DESCRIPTION

Section 2.1 presents a brief history of e-Calc 2’s development. Section 2.2 presents relevant technical considerations. Section 2.3 details the functional logic governing the software. Section
2.4 identifies the required input data and associated usage.

2.1   Developmental History

Minnich and Scotto is the architect of e-Calc – an emissions-calculation software package developed in order to generate air pollutant emission rates from a wide range of fugitive-type, ground-level sources (as well as elevated area sources). This Windows-based, client-server software calculates contaminant emission rates – precise 15-minute-averaged “snapshots” – from these source types. E-Calc is suitable for use with a TDL spectrometer, or any other optical remote sensing (ORS) instrument which generates a path-integrated concentration (PIC). The software can also be used with a rapid-sampling, mobile point-monitoring device, such as a cavity ring-down spectrometer, from which a PIC output can be approximated.

E-Calc is a logical extension of our 2004 PICMET (Path-Integrated Concentration – Meteorology) software, created to rapidly assess compliance with pre-established action levels at off-site receptors (e.g., residences), primarily during hazardous waste site cleanups. The PICMET software displays maximum concentrations at user-specified distances downwind of the emissions source, based on path-integrated measurements and atmospheric stability and transport considerations.

PICMET was employed during active cleanups at former manufactured gas plant (MGP) sites in November 2004, and again during December 2006 and May 2007 as part of a 2½-year applied R&D study for the Gas Technology Institute (Des Plaines, Illinois). Results from this latter study demonstrated superior residential protection when compared to traditional monitoring approaches.

Development work on e-Calc began in 2008. E-Calc was originally created for use with open-path Fourier-transform infrared (FTIR) spectroscopy to help municipal solid waste landfill owners comply with mandated emissions reporting and permitting requirements for methane and other greenhouse gases. Based on AERMOD, the software incorporates the output from the PIC-generating instrument with coincident onsite meteorological data and other information.

In June 2011, we employed e-Calc to support a legal proceeding by measuring emission rates from several process sources at an Alabama pulp-and-paper mill, including a 1-square-kilometer polishing pond. In August 2014, we used it to measure emission rates from the preliminary settling tanks at a large New York City municipal wastewater treatment plant. In September 2015, we participated in an extensive field project for the South Coast Air Quality Management District (SCAQMD), a California governmental agency, in which we used e-Calc to measure emission rates from 16 oil production wells and tanks, 17 gas stations, and two cattle farms, all in the Los Angeles basin.

We have participated in two third-party e-Calc validation studies, results of which were presented at the March 2016 “Air Quality Measurement Methods and Technology Conference,” sponsored jointly by the Air & Waste Management Association (A&WMA) and the U.S. EPA.

First, as part of our project for the SCAQMD (described above), our e-Calc software was validated during a 2-day, controlled-release experiment (October 12-13, 2015). Over the study, propane was released at varying emission rates from a scissors-type lift at a pre-designated height of 3 meters (even though e-Calc was designed for ground-level releases only). Thirteen monitoring events (15- minute-averaged) were performed on Day 1, with an additional seven on Day 2.

Next, under contract to Texas A&M University (San Antonio, Texas), we performed a 2-day, e-Calc validation study (November 4-5, 2015), which involved the controlled release of sulfur hexafluoride (SF6) from ground-level locations simulating a compressor/condensate tank complex (Day 1) and an assembly of gas-gathering pipelines (Day 2).

As mentioned, e-Calc 1 employs the U.S. EPA regulatory version of AERMOD in order to preserve the model’s legal Guideline status. For each monitoring event, the generation of input files requires meteorological data together with emissions-characterization and monitoring configuration data. Dispersion coefficients under this approach (i.e., flux-gradient) are assigned based on wind speed, land-use, solar insolation, and statistical data treatments such as the standard deviations of the horizontal wind direction and vertical wind speed. From this information the friction velocity is determined, which is used to develop the vertical wind-speed profile. The vertical wind-speed profile primarily governs the predicted (back-calculated) emission rate in e-Calc 1 (and e-Calc 2).

The flux-gradient approach currently employed in AERMOD has been extensively evaluated in model-validation studies performed by the U.S. EPA over the years. Similarly, the performance of e-Calc 1 was successfully demonstrated during the two validation studies described above.

The upgraded (second-generation) version of e-Calc (e-Calc 2) was created specifically for this ERA project, primarily to eliminate the need for relatively labor-intensive pre-field tasks. As mentioned, e-Calc 2 employs a more sophisticated means of assigning dispersion coefficients – the eddy- correlation (or covariance) approach. This approach typically requires wind measurements (using sonic anemometry) at two heights above the ground. Covariance statistics, calculated from the lower of these two sensors, are then used to determine the friction velocity. The U.S. EPA is planning to update AERMOD to enable application of the eddy-correlation approach, but has yet to release the software coding for this version.

Additional information about how the vertical wind-speed profile is generated in e-Calc 2 can be found in Section 6.1.1 of this report...

Download PDF to continue reading report, footnotes, figures, tables, and references.

Timothy R. Minnich, President, MS, QEP, is a Meteorologist and Atmospheric Scientist with over 40 years experience in the design and management of a wide range of ambient air and meteorological investigations under CERCLA and the Clean Air Act. He is a recognized technical expert on high-profile legal cases, with assignments involving forensic meteorology and reconstruction of inhalation scenarios in relation to community exposure to hazardous air pollutants (HAP). He is a nationally recognized expert in the application of optical remote sensing (ORS) for hazardous waste site remediation. He has designed and managed more than 25 ORS field investigations and air dispersion model validation studies since the promulgation of U.S.EPA (EPA) Method TO-16 for open-path FTIR (Fourier-transform infrared) spectroscopy in 1988.

©Copyright - All Rights Reserved

DO NOT REPRODUCE WITHOUT WRITTEN PERMISSION BY AUTHOR.