The U.S. Environmental Protection Agency (EPA) is currently reviewing draft revisions to its guidance for estimating emissions from storage tanks as published in Chapter 7.1 of EPA's AP-42 document.1 This guidance presently addresses emissions from storage tanks during routine operations, as well as from the specific non-routine case of floating roof landings. The draft revisions include new guidance for estimating emissions from tank cleaning events, as well as numerous changes to the existing guidance for routine emissions. These revisions would have varying degrees of impact on changes to estimated emissions for most storage tanks.
Other changes under review in the draft revisions to AP-42 Chapter 7.1 include new guidance on the following:
- Estimating flashing emissions;
- Calculating net throughput for working loss from cumulative changes in the liquid level, rather than from pumping volume;
- Estimating emissions from tanks storing hot stocks, accounting for whether the tank is uninsulated, partially insulated, or fully insulated; and
- Estimating emissions from low-pressure tanks.
Revisions to Temperature Equations
Basis for Revisions
Equations in AP-42 Chapter 7.1 calculate temperatures inside a storage tank on the basis of assumed equilibrium with average ambient air temperature, with an additional term to account for the amount of heat from incident solar radiation (“insolation”) absorbed by the tank. The draft revisions to Chapter 7.1 include changes to these equations as well as a change to the manner in which insolation is evaluated.
Significance Level in the Calculations. The storage tank temperature equations were derived from a theoretical energy transfer model that includes numerous parameters, some of which were assigned default values and some of which were retained as variables. For example, angle of the sun, reflectivity of the surrounding ground surfaces, average liquid level, and insulation value of a given floating roof design were all assigned default values, whereas ambient temperature, liquid bulk temperature, average insolation, and reflectivity of the tank exterior surfaces were retained as variables in the final equations.
The resulting equations have significant uncertainty when applied to a specific scenario, due to the use of both default values as well as average values. Default values assigned to certain parameters may not be representative for a particular scenario, and the average value used for a parameter retained as a variable may not represent the actual value when that parameter can vary greatly over time (e.g., ambient temperature). Given the uncertainty in this approach to calculating storage tank temperatures, the prior practice of expressing the coefficients in the equations to two significant figures was deemed inappropriate and potentially misleading, and thus the coefficients in the draft revisions are expressed to only one significant figure.
Inconsistencies in the Prior Equations. Derivation of the temperature equations that have been in AP-42 Chapter 7.1 was documented in API Publication 19.1D.2 Review of this documentation revealed multiple inconsistencies, such as the use of conflicting default values of 0.45 and 2.0 for the tank height-to-diameter ratio. The draft revisions use a consistent default ratio of 0.5. While the dimensions of storage tanks in the petroleum industry vary dramatically, many tanks are 40 or 48 feet tall and 90 to 120 feet in diameter. The draft revisions additionally include a more general form of these equations in which the height-to-diameter ratio is left as a variable rather than being assigned a default value at all.
Prior Simplifications Now Unnecessary. The equations presently in AP-42 Chapter 7.1 were developed prior to the proliferation of desktop computers, when there was a tendency to simplify the calculations by using approximations and substitutions. Given modern computing capabilities, such simplifications are now unnecessary, and the equations in the draft revisions to AP-42 more accurately reflect the theoretical derivations.
For example, the theoretical derivation in API Publication 19.1D used a “gas space” (vapor space) temperature (alternatively labeled Tg or Tv), but this vapor space temperature was not used in the final equations. Instead, to avoid having to calculate this value, the final equations substituted the average liquid surface temperature (Tla) for the vapor space temperature in calculations such as that for vapor density (Wv). The draft revisions include a calculation of the vapor space temperature, Tv, and use of that calculated temperature in the equation for vapor density, Wv.
Evaluation of Absorbed Insolation. The amount of insolation assumed to be absorbed by a tank is a function of the reflectivity of the tank exterior. When solar radiation strikes a surface, a portion of the radiation is reflected and the remainder is absorbed. This absorbed insolation is a source of heat input to the receiving surface. When that receiving surface is a tank shell or roof, the temperature increases inside the tank.
The temperature equations in AP-42 include terms to account for heat input due to absorbed insolation. These terms are a function of the average insolation for the tank location and the solar absorptance of the tank exterior. Paint absorptance factors in AP-42 are based on the color and condition of the tank exterior, with the condition characterized as either “good” or “poor.” The draft revisions re-label these categories of exterior condition as “new” and “aged,” and add a category labeled “average.” Values for absorptance in the average category are simply the average of the values given for the other two conditions.
Impact of Temperature Equations Changes
Fixed-Roof Tanks. For fixed-roof tanks, the draft revisions to the temperature equations generally result in a slightly lower estimate of emissions when assuming the paint condition is “new” (which corresponds to the previous designation of “good”), but a slightly higher estimate when assigning the solar absorptance value for the “average” condition.
For example, consider a white tank that is 60 ft in diameter and 48 ft tall, storing diesel in Port Arthur, TX, with a throughput of 600,000 bbl/yr. Estimated emissions from this tank decrease by 2-4% when applying the draft changes if the “new” paint condition is assigned, but increase by 1-2% if the “average” paint condition is assigned. (The range in the percent change is due to other variables, such as whether the bulk liquid is assumed to be in equilibrium with ambient conditions or is assigned a warmer-than-ambient value.)
External Floating-Roof Tanks. For external floating-roof tanks, the effect of the draft revisions depends on the assumptions regarding the liquid bulk temperature relative to ambient conditions. The trend appears to be a slightly higher emissions estimate when assuming the liquid bulk temperature to be in equilibrium with ambient conditions, but a slightly lower estimate when assigning a measured liquid bulk temperature that is warmer than ambient.
For example, consider a white external floating-roof tank that is 150 ft in diameter and 48 ft tall, storing RVP 7 crude oil in Port Arthur, TX, with a throughput of 3,000,000 bbl/yr. Estimated emissions from this tank increase by 1-3% if calculating the liquid bulk temperature from ambient temperature, but decrease by 3-5% if assigning an elevated value of 80oF as the liquid bulk temperature. (The range in the percent change is due to other variables, such as whether the paint condition is deemed to be “new” or “average.”)
The revisions currently under consideration by EPA to its AP 42 Chapter 7.1 guidance for estimating emissions from storage tanks would have varying degrees of impact on emissions estimated for most storage tanks. Changes to the guidance for estimating emissions from routine tank operations pertain primarily to refinements to the equations for estimating temperatures in a storage tank. The draft changes to these equations will result in higher estimated emissions in some cases, but lower in others. Although the estimated emissions will typically be within ten percent of the previously estimated values, the difference may be greater in certain scenarios.
Guidance is also being considered for scenarios that had not previously been addressed in AP-42 Chapter 7.1, such as estimating emissions from the cleaning of storage tanks and adjustment to the equations for estimating emissions from routine operations in the case of partially or fully insulated tanks.
Changes to emission factors always raise questions as to when the new factors must be used for both permits and emissions inventories. Because these programs are administered at the state level, implementation may vary depending on facility location.
1 U.S. Environmental Protection Agency, 7.1 “Organic Liquid Storage Tanks,” in Compilation of Air Pollutant Emission Factors, USEPA Report No. AP-42, November 2006.
2 American Petroleum Institute, Publication Chapter 19.1D, “Documentation File for API Manual of Petroleum Measurement Standards Chapter 19.1 - Evaporative Loss From Fixed Roof Tanks,” First Edition, March 1993.