Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.
- Gasoline evaporative losses
- Sources of gasoline evaporative losses
- Data sources
- Results and estimates: evaporative losses
- Concluding remarks
The Survey of Industrial Processes (SIP) was conducted with one primary objective in mind. To test a methodology for collecting data on operational activities and industrial processes that would facilitate modelling/imputation of pollutant releases from small and medium-sized enterprises across Canada.
As a proof of concept, the SIP pilot focused on the operational activities and industrial processes used in retail gasoline outlets across Canada. Retail gasoline outlets and their associated gasoline vapour emissions were selected in cooperation with Environment Canada as an important data gap. The survey population included retail gas stations as well as marinas with gas docks. It comprised establishments primarily engaged in retailing gasoline fuel whether or not the outlet was operated in conjunction with a convenience store, repair garage, restaurant or other type of operation. Diesel-only outlets and card-locks were excluded from the population. The survey employed standard statistical methods in the collection of data including the design of the survey frame, sample selection, and statistical edits, imputations, and weighting methodologies. 1 Data collected include quantities of gasoline sold, gasoline truck delivery frequencies, number of gasoline storage tanks, number of fuel dispensers, age of site, and other related activities and processes that depict this particular sector of the Canadian economy. 2
This report presents the modelling method used for the estimation of gasoline evaporative losses derived from the integration of the SIP survey data with a set of mathematical models 3 , 4 , 5 . The results presented here should be considered as a proof of concept only and are not intended as official statistical estimates.
The survey results can complement the area source (AS) component of Environment Canada's Air Pollutant Emissions Inventory while the National Pollutant Release Inventory (NPRI) accounts for point sources (PS). The NPRI collates data from a number of sources and presents estimates on the emissions of a number of air pollutants including volatile organic compounds that are the components of gasoline emitted to air. One part of the NPRI collects pollutant release data from establishments/facilities with emissions above given thresholds. The threshold applied to the NPRI limits its statistical coverage. It restricts the inventory to large emitters, leaving out medium and small emitters. These small and medium emitters together could add significantly to Canadian emissions of some pollutants. In light of the above, the SIP was conducted to address some of the data gaps present in the Canadian air emission inventories. Also, due to its bottom-up approach of collecting industry and process data though a survey, the SIP is potentially able to generate pollution estimates at levels of geography down to large cities.
This report is organized as follows. The following section describes gasoline evaporative losses in general and mentions some of the efforts to measure them elsewhere. The sources of evaporative losses from retail gasoline outlets are described next, with specific emphasis on Canadian outlets. The subsequent section presents the data sources used to generate the illustrative gasoline evaporative loss estimates from retail gasoline outlets across Canada including the initial survey results. The estimates themselves are presented in the ensuing section. An appendix outlines the mathematical models, uncertainties, and the assumptions used to generate the estimates.
2 Gasoline evaporative losses
Gasoline is a mixture of volatile organic compounds (VOC), mostly hydrocarbons that can freely evaporate at ambient temperatures. Retail gasoline outlets contribute to overall VOC emissions across Canada through fuel-related activities carried out on a daily basis. Gasoline contains compounds such as benzene, toluene, and ethylbenze, some of which have adverse impacts on human health. Benzene, for example is a carcinogen that is present in gasoline. 6 According to the United States Environmental Protection Agency (USEPA), the concentrations of benzene, ethylbenzene, toluene, and xylene around homes within 200 meters of retail gas stations are higher than urban background levels. 7 Gasoline vapour, once released to the atmosphere, combines with nitrogen oxides in the presence of sunlight and forms ozone. Ground-level ozone is a component of smog that aggravates respiratory illnesses. 8
Estimation of evaporative losses from the gasoline distribution network, and in particular from retail gasoline outlets, has been the subject of several studies. For example, the California Energy Commission collects information on all retail fuelling stations located in California through the California Annual Retail Fuel Establishment Form which needs to be filled out by these facilities as a mandatory report. 9 Environment Australia estimates the aggregated emissions from service stations using information collected in the Australian National Pollutant Inventory (NPI). 10 According to the European Emission Inventory Guidebook, the contribution of the gasoline distribution sector to the total man-made non-methane volatile organic compounds emissions ranged from 1.5 % to 6.7%. The reported values apply to all 28 European countries that participate in the inventory. 11
At the gas station/outlet level, some studies focused on empirical/experimental data while others used theoretical models based on established thermodynamic and gas-liquid equilibrium principles. Mathematical models to correlate the evaporation of gasoline and its components with other relevant factors have been developed and established. 12 , 13 , 14 , 15 The method used to estimate evaporative losses in this report builds upon these earlier studies.
3 Sources of gasoline evaporative losses
Evaporative losses from retail gasoline outlets can be classified into two categories. The first category includes those attributed to the processes and equipment used by the retail outlet to supply the gasoline. The second category includes losses attributed to vehicles being filled (vehicle tank refuelling loss). Following this classification, the sources of gasoline evaporative losses from retail outlets can be detailed as follows:
3.1 Losses from standard operations of a retail gasoline outlet
3.1.1 Storage tank losses
Working loss in gasoline storage tanks is due to the combined effects of gasoline delivery to the storage tanks (the filling of storage tanks) and the emptying operation (the pumping of gasoline from storage tanks to the dispensers/gas pumps). During the filling of storage tanks, fuel vapours are released to the atmosphere due to the increased liquid level in the tank pushing up and reducing the vapour space in the tank. This space is also called outage space or headspace.
The vapour is consequently compressed in the tank, forcing the air-vapour mixture out through a vent pipe. If the tank is equipped with a pressure/vacuum valve on its vent pipe, vapours are released only when the pressure inside the tank exceeds the valve relief pressure. In the absence of such valves, as is the case in most of Canada, 16 any increase of tank pressure above atmospheric levels will release the vapours through the open vent pipe. The absence of such pressure/vacuum (P/V) valves on the storage tank vents of Canadian retail gasoline outlets was traditionally based on the consensus that P/V valves may freeze during the winter season, risking the implosion of storage tanks. This has freed retail gasoline outlets from the mandate of installing P/V valves and left the choice to do so to their own discretion.
In some parts of Canada, tanker truck drivers who deliver gasoline to retail gasoline outlets are required by law to attach a second hose to the storage tank while refilling it to capture part of the air-vapour mixture that would have otherwise escaped from the vent pipe. This is called vapour balancing. The effectiveness of this particular "second-hose" vapour recovery method (that is, vapour balancing) in the absence of a P/V valve on the vent pipe of a tank is uncertain. Although some reports claim 90% recovery, these reports have mostly assumed the presence of a P/V valve or some sort of an orifice restriction during gasoline tanker truck deliveries. 17 , 18
Two methods are commonly used in filling storage tanks in retail gasoline outlets: splash filling and submerged filling. Significant liquid turbulence and vapour/liquid contact occur during a splash filling operation, resulting in high levels of vapour generation and subsequently high evaporative losses through the vent pipe. In submerged filling, the drop tube extends close to the base of the storage tank with fresh fuel being dispensed below the original liquid surface level. Liquid turbulence is thus controlled, resulting in much lower vapour generation and less evaporative losses than in the case of splash filling method.
According to the SIP results, there are some 29,000 retail gasoline storage tanks across Canada and it is reasonable to assume that not all of them are equipped with submerged fill pipes. How many are so equipped is unknown, since respondents to the SIP were not asked for this information. None of the outlet operators could answer this particular question during the testing of the draft questionnaire, so the question was not included on the final questionnaire. For this reason, an uncertainty algorithm that includes both submerged and splash filling was applied in calculation of the evaporative loss estimates.
Since dirt particles tend to settle at the bottom of storage tanks, the length of submerged tubes is such that they reach approximately 6 to 12 inches above the bottom of the tank in order to prevent settled particles from being agitated. It is assumed here that at the outset of any filling operation some splash filling occurs, while gradually the filling operation becomes fully submerged.
The estimation method used in this study uses a ratio to account for the application of such hybrid filling (splash and submerged). A ratio was estimated based on the assumption that the fill pipe is 12 inches above the bottom of the tank and the level of gasoline in an "empty" tank is always 6 inches from the base; that is, no storage tank is completely emptied before its next filling. This provides a 6-inch empty space that is always associated with splash filling.
Gasoline pumping activities during the refuelling of vehicles also contribute to the working loss in storage tanks. This occurs due to air being drawn through the vent pipe into the storage tanks as a result of the decrease in storage tank liquid level while pumping out gasoline. Once again this action is more pronounced in storage tanks that are not equipped with a pressure/vacuum valve on their vents as a control measure. Any additional entrained air becomes saturated with gasoline vapours and subsequently expands. The expanded air-vapour mixture exceeds the vapour holding capacity of the storage tank and is consequently expelled back to the atmosphere through the same vent pipe.
Gasoline evaporation due to working loss from retail gasoline outlets is estimated based on a USEPA model described in Appendix I.
Breathing loss (also known as standing loss) is due to the release of gasoline vapour to the ambient air due to the expansion and contraction of vapour inside storage tanks. The breathing loss results from changes in the temperature and pressure inside and outside of the storage tank. Breathing loss occurs irrespective of changes in the liquid level inside the tank.
Gasoline storage tanks are exposed to changes in temperature, pressure and, in the case of aboveground tanks, to solar radiation. As a result, the mixture of trapped air and gasoline vapour inside the tank undergoes passive heating and cooling cycles. This causes expansion and contraction of the vapour mixture, resulting in the release of vapour to the surrounding environment. In the absence of a control measure such as a P/V valve, this "breathing process" continues as long as there is some gasoline liquid inside an open-vent tank.
Since changes in temperature and pressure are less pronounced for underground tanks due to the protective and insulating effects of soil, the USEPA does not associate breathing loss with underground storage tanks. This is, in part, justifiable for several reasons. One reason is that most cities in the USA are in climates where the difference between the ambient air and the underground temperature is less pronounced than in most Canadian cities. Also underground tanks in the USA are generally equipped with vent pipes that have P/V valves with pre-set pressure and vacuum settings to control the breathing activity. As mentioned earlier, P/V valves are not common in Canada. 19 Instead, Canadian storage tanks generally employ valve-free vent pipes that are open to the atmosphere. In light of the above, unlike in the USA, breathing losses in Canadian underground gasoline tanks should be included in the evaporative loss estimates regardless of the partially insulating effects of being underground. Consequently, the estimates of evaporative losses reported here include breathing losses for both underground and aboveground storage tanks.
In the case of underground tanks, the effect of solar radiation was removed from the calculation and only the impact of temperature was included. The difference between air temperature inside an underground tank and that of the ambient air temperature outside was estimated to be 8 degrees Celsius during the cold season 20 and 12 degrees Celsius during the warm season. 21 , 22 These parameters were used in estimating breathing losses in underground gasoline storage tanks across Canada.
Gasoline evaporation from retail gasoline outlets due to the breathing loss is estimated based on a USEPA model described in Appendix I.
3.1.2 Residual losses
Operations and activities that involve opening the lids of gasoline storage tanks trigger the release of gasoline vapours to the atmosphere. Gasoline vapours accumulate in the headspace above the liquid gasoline level of tanks. This activity, though a minor contributor, is nonetheless an extra source of evaporative losses in retail gasoline outlets. Lids of gasoline tanks are regularly opened by the operators of the outlet to check the level of gasoline and for the presence of water in the tanks. Storage tank lids are also regularly opened by the gasoline delivery personnel during their filling operations. However, since Canadian gasoline storage tanks are generally open-vented tanks (that is, no P/V valves on the vents), any evaporative losses from opening the tank lids are minute. To that end, evaporative losses from this particular activity were considered negligible for the purposes here.
Monitoring also involves inserting a wooden dipstick into the storage tanks to measure the gasoline liquid levels and to check for the presence of water. The measurement of gasoline level and water monitoring in storage tanks is carried out on a regular basis. The measurement of gasoline level is conducted by both the operator of the retail outlet and the gasoline delivery personnel, independently. It involves the immersion of a graduated wooden stick into the liquid inside the tanks to determine the level of gasoline. In addition, operators use the wooden stick to check for the presence of water at the base of the tank by placing a special paste at the tip of the stick and observing any change in the colour of the paste.
Monitoring losses from dipsticks occur due to the evaporation of gasoline that has been adsorbed on the wooden surface of the dipsticks. These dipsticks are subsequently placed in the open air, causing the evaporation of adsorbed gasoline on the surfaces of these wooden sticks into the atmosphere. Fixed electronic level-monitoring devices are also available and are commonly used in many retail gasoline outlets across Canada, although the use of dipsticks remains common by gasoline delivery personnel verifying their deliveries. 23 Based on the activity data collected from the SIP and a laboratory experiment, the reported evaporative loss estimates have incorporated an empirical model to estimate losses from the use of dipsticks.
Gasoline evaporation from retail gasoline outlets due to monitoring losses related to the application of dipsticks are estimated based on an empirical model described in Appendix I.
A residual loss is any uncontrolled leak/spill that occurs from dispensers. Part of this residual loss is from nozzle spills. Since absorbents are mainly used by retail gasoline operators to clean up after this particular type of spill, it is assumed that the amount of such spills is correlated with the quantity of absorbents used. Note that at marinas with gas docks absorbent use is limited and as such any calculated residual loss at these facilities can be almost entirely attributed to leaks/spills at the dispenser.
Gasoline evaporation from retail gasoline outlets due to the residual loss is estimated based on USEPA emission factors and on an empirical model based on the capacity of absorbents to hold gasoline and the amount of absorbents used at the outlets as reported in the SIP. Both the emission factors and the empirical model are described in Appendix I.
3.2 Losses from the refuelling of vehicles
Refuelling emissions occur when vapour from the headspace of a vehicle fuel tank is displaced by the liquid gasoline that is dispensed into the fuel tank. The volume of displaced vapour during the refuelling operations is equal to the volume of gasoline dispensed into the vehicle fuel tank, plus the entrapped droplets of liquid fuel as a result of splashing and turbulence during filling which are subsequently released as vapour. The quantity of displaced vapours depends on the temperature of gasoline in the vehicle fuel tank, temperature of dispensed gasoline, gasoline Reid vapour pressure (RVP), and the dispensed volume of gasoline. The volume of vapour released during refuelling also depends on the vapour recovery method used. In the USA, many dispensers at retail gasoline outlets incorporate vacuum-based nozzles as a vapour recovery method. In the USA they also rely on the on-board vapour recovery (OBVR) system installed inside vehicles. In Canada, only the latter has been applied as the vapour recovery system during vehicle refuelling. 24 In fact, the combined use of the two methods could defeat the purpose since the former method counters the effectiveness of the latter. 25 In Canada, OBVRs have been installed in all newly manufactured vehicles since 1998. OBVRs are up to 98% efficient in capturing evaporative losses during a refuelling operation. 26 However, to account for operational and efficiency degradation of the OBVR systems, a recovery efficiency range of 90±5% was set for OBVRs for use in calculating the estimates reported here.
For evaporative loss estimates, a percentage of OBVR penetration based on the age of Canadian vehicles was used. Based on data from Environment Canada, 27 this percentage was estimated at 70±5% for the year 2009.
Evaporative losses due to the refuelling of vehicles are estimated based on a USEPA model described in Appendix I.
4 Data sources
4.1 Survey data
The SIP pilot was the main source of the data used in this study. 28 For the 2009 reference period, the SIP pilot survey covered all retail gasoline outlets, including marinas with gas docks, across Canada. All data referenced in this section are from the 2009 SIP pilot unless otherwise noted.
According to a National Retail Petroleum Site Census published by MJ Ervin and Associates, two decades ago there were over 21,000 retail gasoline outlets across Canada. 29 In 2009, there were fewer than 11,300 gasoline outlets in operation.
In 2009, approximately 40.7 billion litres of gasoline were sold at retail outlets across Canada (Table 1). Based on an average delivery of approximately 35,000 litres, it is estimated that over 1.1 million gasoline deliveries were needed to support the gasoline consumption of vehicles in Canada in 2009.
4.2 Spatial data
A map representing the spatial distribution of gasoline outlets across Canada was created by using the postal code of each outlet to estimate its latitude and longitude. In major cities, this allocation method rendered a spatial distribution with a good accuracy. In less populated areas where postal codes cover larger areas, this method, though less accurate, was acceptable for the purposes here (See Map 1).
4.3 Meteorological data
Based on proximity to a meteorological station, retail outlet/site specific ambient temperatures (average, maximum, and minimum for warm and cold seasons, 2009, respectively) were obtained from Environment Canada. 30 Solar radiation values (average peak summer and winter) were extracted from the National Atlas of Canada 31 and digitized/mapped over the locations of gasoline retail outlets across Canada using geographical information system technology.
4.4 Reid vapour pressure data
Reid vapour pressure for gasoline deliveries during cold and warm seasons were supplied by Environment Canada 32 and values were mapped and matched to respective retail gasoline outlet locations using geographical information system technology.
5 Results and estimates: evaporative losses
Data from the SIP were used along with the models described in the appendix I to calculate an estimate for gasoline evaporative losses for each retail gasoline outlet that reported to the SIP. The estimates addressed evaporative losses associated with on-site gasoline truck deliveries, storage tanks, vehicle refuelling, and other activities/processes and control measures related to retail gasoline outlets. Each individual evaporative loss estimate was weighted so that all of the individual estimates taken together represented the entire population of retail gasoline outlets across Canada. Sample weights were based on the statistical methodology employed in the SIP. 33
5.1 Gasoline outlets across Canada
In 2009, approximately 58.3 million litres of gasoline (in liquid equivalents) was evaporated from some 11,200 retail gasoline outlets across Canada. This is equivalent to the contents of one full tanker truck evaporating approximately every 8 hours.
5.1.1 On-road retail gasoline outlets
The following table presents a breakdown of gasoline evaporative losses from on-road gasoline outlets alone. Two-thirds of evaporative loss was due to the refilling of the 29,000 retail gasoline storage tanks that were in operation during 2009. The remaining third was from the refuelling of gasoline vehicles themselves.
5.1.2 Marinas with gas docks
In 2009, gasoline evaporative losses from all Canadian marinas with gas docks totalled 157,000 litres. This accounted for 0.3% of total national retail gasoline evaporative losses.
6 Concluding remarks
The SIP was designed as a pilot test of the use of economic and operational data collected through a statistical survey in the estimation of the release of certain criteria air contaminants (CACs) from small and medium-sized enterprises (SMEs) within a given sector of the Canadian economy.
As a proof of concept, the SIP pilot focused on the operational activities and industrial processes used in retail gasoline outlets across Canada. Retail gasoline outlets and their associated evaporative losses were selected in cooperation with Environment Canada as an important data gap.
The resulting estimates of gasoline evaporative losses compare well with Environment Canada's estimates once uncertainties are included. Environment Canada has estimated that the VOC emissions from the retail gasoline sector were approximately 68 million litres 34 in 2009 with a qualitative uncertainty of 25%. 35 This estimate overlaps statistically with that based on the SIP results (58 million litres with a quantitative uncertainty of 2.5% at 90% confidence level).
Although the two estimates were produced using different models in terms of scale and approach, the overlap in the two estimates is promising. The SIP, using a detailed bottom-up approach to estimate evaporative losses, has statistically validated the aggregated top-down approach of Environment Canada. The latter used total national gasoline sales with relevant emission factors and temporal adjustments.
The advantage of the SIP is in its statistical basis, with a broad population coverage, statistical weights, and quantifiable uncertainties. The SIP provides a rich pool of microdata from which pollution estimates can be generated at a much finer level of geography than is currently possible in the national pollutant inventory; for example, at the large city level. This in turn facilitates the generation of pollution concentration gradients across regions; that is, those with more stringent gasoline delivery regulations versus those with less stringent regulations to help compare the effects of regulations as they vary from location to location or province to province.
In addition, the detailed data collected by the SIP in terms of economic and operational activities, industrial processes and equipment offers the ability to generate different emission scenarios by changing assumptions around, for example, gasoline outlet operational practices or regulations.
It is an open question whether a similar outcome could be achieved for other industries and other pollutants. Further work would be required to identify existing emission models and/or develop new ones and additional industrial process surveys would be required to collect the necessary operational and process data. At the moment, this remains work for the future.