NPRI List of Reasons for Yearly Change

1. Introduction

In 2002, Environment Canada (EC) added the Criteria Air Contaminants (CAC) to the list of the substances reported to the National Pollutant Release Inventory (NPRI). A few years later, a sub-group (SG) of the 2006 NPRI Multi-stakeholder Work Group on Substances (WG) was created to review the data quality issues and data gaps surrounding reporting of CAC to the NPRI for the compilation of comprehensive air emissions inventories. The final CAC SGs report was discussed by the WG at its February 2007 meeting. At that same meeting, the WG decided to launched, two new CAC SGs : CAC-Volatile Organic Compounds SG (CAC-VOC SG) and CAC-Threshold SG (CAC-TH SG). These two new SGs were mandated to address and resolve 31 remaining CAC issues, and make recommendations to the WG at the end of their respective mandates.

Out of the 32 original CAC SGs recommendations, one recommendation has not yet been addressed by the two current SGs: large variation in the NPRI emissions reported for some facilities; specifically, the lack of information on the reasons for variation in the emissions from year to year.

Originally, EC proposed to resolve this issue by adding a new field for facilities to provide explanations on the yearly variations of each CAC emission that was greater than 20% for a specific year. However, upon further analysis and a comprehensive review of emission change comments, the list of reasons provided to explain emission changes, or Reasons for Change (RFC), can provide a focus for possible improvements.

2. Current Data Overview

2.1. CAC Emissions

After the analysis were completed, it was found that the overall total emissions did not vary greatly (standard deviation of 1.7% of total emissions average) and the largest contributors to CAC emissions are Sulphur dioxide (43.3%), Carbon monoxide (23.3%) and Oxides of nitrides (18.4%). A summary of the yearly CAC emissions is provided in Table 1.

Table 1: Annual Release to Air of CAC (in tonnes)
Substance20022003200420052006
Ammonia17,58817,72319,56620,20221,624
Carbon monoxide996,930962,8141,067,4101,026,2851,047,840
Oxides of nitrogen67,4928846,901850,428841,330825,690
Total PM246,800191,311176,671174,990198,514
PM10107,013109,970104,514101,963110,477
PM2.559,57664,38662,15761,72162,582
Sulphur dioxide197,85711,944,7061,952,8771,851,3741,773,639
VOC275,731270,000269,233262,110271,048
Total4,357,1364,407,8124,502,8584,339,9744,311,413

Table 2 shows the total variance between two consecutive years by providing the summation of the absolute differences of facilities' yearly emissions. This demonstrates the full volume of change for CAC releases, i.e. the total accounts for both negative and positive changes.

Table 2: Total Variation of Annual Release to Air of CAC (in tonnes)
Substance2002-20032003-20042004-20052005-2006
Ammonia3,1593,7424,1083,214
Carbon monoxide345,532319,496330,609233,639
Oxides of nitrogen213,896178,242150,297176,131
Total PM107,03249,59443,36159,089
PM1039,52727,01626,76130,739
PM2.521,55715,58118,13720,310
Sulphur dioxide292,811290,109327,745272,422
VOC84,74687,12359,65972,770
Total1,108,260970,903960,678868,314

The following observations can be drawn from the CAC emissions data:

  • Absolute total emissions variations have decreased over the years.
  • Carbon monoxide (31.5%), Sulphur dioxide (30.3%) and Oxides of nitrogen (18.4%) provide the greats variations.

2.2. Reasons for Change (RFC)

During yearly emissions reporting, facilities are prompted to provide more information on their RFC from the previous year's emissions. A list of reasons is provided (multiple selections are possible), along with a text box for additional comments. Table 2.3 presents the current list of RFC in quantities release from previous year, by release type. The text is from OWNERS as it appears.

Table 3: Reasons for Change
Data FieldNon-CAC ReleasesCAC ReleasesDisposalsRecycling
AChanges in production levelsChanges in production levelsChanges in production levelsChanges in production levels
BChanges in estimation methodsChanges in estimation methodsChanges in estimation methodsChanges in estimation methods
CPollution prevention activitiesPollution prevention activitiesPollution prevention activitiesPollution prevention activities
DChanges in on-site treatmentChanges in on-site treatmentChanges in on-site treatmentChanges in on-site treatment
EChanges in off-site transfers for disposal  Changes in disposals
FChanges in off-site transfers for recycling Changes in off-site transfers for recycling 
GOther (specify in comments field)Other (specify in comments field)Other (specify in comments)Other (specify in comments)
HNo significant change (i.e. <10%) or no changeNo significant change (i.e. <10%) or no changeNo significant change (i.e. <10%) or no changeNo significant change (i.e. <10%) or no change
INot applicable (first year reporting this substance)Not applicable (first year reporting this substance)Not applicable (first year reporting this substance)Not applicable (first year reporting this substance)

The following observations can be drawn:

  • "Production" and "Not Significant" RFC account for the majority of emission changes, average of 27.7% and 28.6%.
  • "On-site Treatment" and "Pollution Prevention Activities"RFC account for a small amount of emission changes, average of 0.4% and 1.3%.
  • The use of "Other" as a RFC has increased over the years.
  • "Not Applicable (first year reporting)" decreased dramatically after the first couple years.
  • "Estimation Methods" RFC has averaged around 10%.

3. Data Analysis

3.1. Quantify Reasons for Change

Yearly variations, shown in Table 2, are the result of all the facilities' variations. Those variations are clarified by the facilities, as shown in Section 2.2. Therefore, correlating these two datasets (by facility and CAC substance) will provide the yearly change of a facility associated to a particular RFC.

Ordering facilities by their percent contribution to the total yearly variance and calculating the cumulative sum of their percents, the progressive representation of each RFC can be demonstrated for any reporting year.

Note: In the cases where multiple RFC were selected, the emission change amount was associated to all reasons selected; therefore, the total of individual emission changes will be greater than the total yearly variance.

Observations that can be drawn:

  • "Production" and "Other" RFC account for the majority of the total yearly variations (47% and 40%, resp.), in both reason and emission quantity.
  • "Not Significant" RFC was selected the most times, but is only 10% of the total emission variance.
  • "Estimation Method" RFC accounts for 20% of the total yearly variations and is selected some of the time.
  • "On-site Treatment" and "Not Applicable" RFC do not contribute any variations. "Not Applicable" quantity will be different for earlier years.

To properly address the lack of information on the RFC in emissions from year to year, the highest contributors to yearly variances (i.e. "Production" and "Other") are the initial subjects for investigation. At the outset, it can be seen that these terms are vague and require further development. By examining the additional comments provided during release submissions, which offer insight into the selected RFC, a better suiting list of RFC can be established. This is done in the following section.

3.2. Comprehensive Analysis of Change Comments

By examining the additional comments provided when facilities report emissions, it can be determined if the current options for RFC are relevant and if alternative options would be beneficial. To evaluate the comments, a random sample of comments was extracted (using Excel's RAND function). The data sample included approximately 15% of each release type's comments: CAC Release (852), non-CAC Release (578), Disposal (451) and Recycling (290). English, French and Bilingual comments were selected for the analysis. A total of 2,171 comments were analysed or 15.2% of the total 14,325 comments.

The sample comments were separated by release type and categorized into the appropriate RFC. In some cases, comments could fall into more than one of the categories so, they were counted twice (count total = 2,369). During analysis, categories were modified to best represent the overall impression of RFC type. Table 4 gives the count and frequency (count percent of total sample size) of the final RFC categories.

Table 4: Results of Comments Sample Analysis
RFC CategoryNon-CAC ReleaseCAC ReleaseDisposalRecyclingTotal
#%#%#%#%#%
Changes in quantity and/or composition of combusted materials294.7%328 34.4%12 2.5%0 0.0%369 15.6%
New method of estimation87 14.0%167 17.5%31 6.4%8 2.6%293 12.4%
Activities/events of a non-annual nature79 12.7%106 11.1%66 13.6%27 8.8%278 11.7%
Other528.3%434.5%64 13.2%36 11.7%195 8.2%
No release457.2%       187 7.9%
No disposal    74 15.3%  
No recycling      68 22.1%
Changes in process and/or raw/ produced/processed materials90 14.4%333.5%71 14.6%43 14.0%237 10.0%
Calculation details68 10.9%879.1%15 3.1%18 5.8%188 7.9%
Process/production/ source/substance use description335.3%171.8%53 10.9%26 8.4%129 5.4%
Production change details477.5%444.6%26 5.4%20 6.5%137 5.8%
Explanation of normal variation243.9%181.9%40 8.2%42 13.6%124 5.2%
Pollution prevention details193.0%293.0%4 0.8%11 3.6%63 2.7%
Identical values explanation81.3%353.7%1 0.2%0 0.0%44 1.9%
Error in previous and/or current report(s)142.2%242.5%3 0.6%3 1.0%44 1.9%
Changes in composition of materials released/ transferred213.4%  22 4.5%4 1.3%47 2.0%
Not reported in previous year71.1%222.3%3 0.6%2 0.6%34 1.4%
TOTAL623100%953100%485 100%308 100% 2,369 100%

Further analysis of the data obtained in Table 4 can be summarized as such.

  • When examined by individual substance types/activities, the need for modifications to the list of RFC in OWNERS is even more evident. For example, almost 34.4% of CAC release comments are related to changes in combustion, while none of the recycling comments in the sample are related to combustion. It may be possible that an option for combustion-related changes only needs to be added for CAC releases and not the other three types.
  • The frequency of comments related to activities of a non-annual nature is 11.7% over the whole sample. However, this is as high as 13.6% for disposal comments and as low as 8.8% for recycling comments.
  • The frequency of comments related to changes in process/materials is 10.0% over the whole sample. However, this is as high as 14.6% for disposal comments, 14,4% for non-CACrelease comments and 14.0% for recycling release comments and as low as 3.5% for CAC release comments.
  • Overall, 7.9% of the comments related to the substance not being released/disposed/recycled. However, since this is as a percentage of all substance/activity types and does not apply toCAC releases, the overall frequency is artificially low. When looked at individually, the no release/disposal/recycling category occurs between 7.2% and 22.1% of the time in the sample. The same applies to changes in composition of materials released/transferred, with an overall frequency of 2.0%, compared to a range of 1.3% for recycling comments to 3.4% and 4.5% for non-CAC release comments and disposal comments, respectively.

4. EC Proposal

First, based on the findings of the study of comments provided when reasons for yearly change are reported, Table 5 provides the list of the proposed fields for options of RFC. To further enhance reporting abilities, sub-data fields are formed to provide more fields to query.

Table 5: Proposed Reasons for Change
Data FieldNon-CAC ReleasesCAC ReleasesDisposalsRecycling
A Changes in production levels; does not include weather-related changes in combustion, which are captured under "Changes in combustion". Please provide details in comment field.
 A.1Increase in production level.
 A.2Decrease in production level.
B Changes in estimation methods. In the comment field, please provide details of the old and new EF and/or methods used.
 B.1New emissions factors.
 B.2New calculation method.
 B.3New source testing results.
 B.4Sources added/removed.
C Pollution prevention activities.
D     
E     
F     
G Other. Please specify details in comment field.
H     
I Not applicable (no report for this substance in the previous year).
J Changes in process and/or production process(es); does not include combustion process/material changes, which are captured under "Changes in combustion".
 J.1Quantity change in raw/processed/produced material(s).
 J.2Type change in raw/processed/produced material(s).
 J.3Composition change in raw/processed/produced material(s).
 J.4Addition/removal of production line(s).
 J.5Installation/decommissioning of equipment.
K Changes in combustion.
 K.1Increase in combustion quantity.
 K.2Decrease in combustion quantity.
 K.3Characteristics change in combusted material(s) (e.g. flared gas, wood waste, incineration).
 K.4Installation/decommissioning of combustion equipment (e.g. incinerators, fossil-fuel compressors and boilers).
L Substance was not released. Substance was not disposed of on-site or off-site.Substance was not transferred for off-site recycling.
M No change (based on identical year-to-year unit capacities/production levels).
N Significant activity(ies)/event(s) of a non-annual nature (spill/leak, planned/unplanned temporary/permanent remediation or catalyst removal/regeneration). Does not include facility-wide shutdown activity(ies).

Second, the following conditions should be used to ensure the best feedback is provided for quality analysis:

  • All reasons for change fields are to be reset each year to ensure an active selection (i.e. no pre-filling).
  • "Other" and "Estimation Method" RFC would require input in the comment field (i.e. users cannot continue through the submission process until additional comments are provided). The otherRFC would have comments requested but not required.
  • "Significant activity(ies)/event(s)" RFC would have optional start and end date fields that will be used to identify the time the non-annual activity/event took place.
  • A general note should be included for all RFC: "Please provide as much detail as possible in the comment field. This will help to avoid the need for NPRI staff to contact you during the annual NPRI data review for clarification of reported data."

This proposal addresses the lack of information in yearly variations by providing a relevant and comprehensive list of RFC that will minimize the use of "Other" as a RFC and elaborate on the "Production" RFC. Also, any RFC in the full range of options readily used in a query and will minimize the need to analyze the text comments to create meaningful trend analysis reports. However, forcing additional comments for "Other" and "Estimation Method" RFC will provide an opportunity for further quality analysis if needed.

EC is seeking comments on this proposal, and would like to hear from the stakeholder community.

May 27, 2008