GCI TECH NOTES ©


Volume 1, Number 8
A Gossman Consulting, Inc. Publication
August 1995

Organic Vapor Control - EPA's Subpart CC Regulation

David Constans

Introduction

The EPA had promulgated additional organic vapor control amendments to 40 CFR 264 and 265. These controls, designated as Subpart CC, were promulgated for an effective date of June 5, 1995. The EPA has since extended the compliance date to December 6, 1995. To comply with the regulations facility operators and waste generators previously exempt under the "less than 90 day storage" regulations must insert into the facility record an implementation schedule for achieving the Subpart CC requirements. The installation and proper operation of the organic vapor control equipment must be completed by December 8, 1997.

Subpart CC is an extension of Subpart AA and BB as well as modifying these subparts. The technology involved in complying with subpart CC is well established and need not be technically demanding in its design or operations. This issue of "Tech Notes" discusses the design of these closed-vent systems and organic vapor control devices.

Design Considerations

Tank systems generally must conform to the National Fire Protection Association Code (NFPA30) or some other accepted code. These codes limit a number of the parameters involved in closed-vent system design, such as the maximum allowable internal pressure in an atmospheric tank and the minimum size for tank vents. A careful reading of the applicable code before designing a closed vent system is recommended.

These systems are certainly nothing new to facilities handling organic hazardous wastes. They are required by Subpart BB and most State air permits. The systems are simply nothing more than gas vent lines connecting each tank into a header. This allows organic vapor contaminated gas displaced from the receiving tanks, during a liquid transfer, to go to the tanks being transferred. The header is subsequently connected to an organic vapor control device. In some state air permits, or localities in non-attainment zones, transport containers (tank trailers, rail cars, even large totes) are required to be connected into the system during transfer. Configured in this manner the organic vapor control device need only control vapors forced out of the system due to normal daily heating and cooling of the gas in the vapor space in each tank (normally called "breathing losses").

It has been reported to GCI that EPA inspectors interpret Subpart CC to consider the vapors vented back into a transport container as a non-compliant activity. No definitive explanation has been given for this interpretation. One would assume that the EPA views the vapors leaving the facility in the container as exiting the system other then through the control device, hence a violation. A decidedly narrow view of organic vapor control.

As stated earlier, many states (and/or localities) require the transport container to be connected to the system during transfers. This is eminently logical as consistent use of these vapor vent back systems suppress the production of organic vapors. No transport container can be totally emptied. There will always be a film of liquid or a heel of material in the container. This is especially the case with wastes that contain solids. Consequently, if vapor laden gases are not vented back to the transport container, this film or heel will evolve vapors until the gas in the container is saturated (or all of the liquid evaporated). Implementing EPA's narrow interpretation of Subpart CC would have the net effect of producing more organic vapors. The vapors displaced during the transfer, since they could not travel back to the transport container, would vent through the organic vapor control device. Meanwhile the transport container would be allowed to draw air or nitrogen which in turn would become saturated. The overall effect is a 100% increase in gross emissions (before treatment in an organic vapor control device) and a possible 5% increase in organic vapor emission. (Based on the EPA mandated requirement of a 95% reduction in organic emissions from the control device.) Clearly this interpretation neither minimizes waste production (organic vapor laden gases) nor reduces emissions. Although this would appear to be a side issue in the design and operation of a vapor control system it is an essential issue in the control device's vapor loading calculation, as shown in the following example.

Breathing losses can be calculated by determining the maximum amount of thermally expanded gas (i.e. due to the heating of the tankage by the sun) that would be produced each day. Assume that there was a 50 F degree change in temperature (70 - 120 F) over a 6 hour period for a tankage vapor space of 250,000 gallons. The organic vapor control system would handle an average 9 CFM of organic laden vapor. Unloading a 5,600 gallon tank trailer in 45 minutes would add an additional 16 CFM of burden, almost tripling the design requirements. The tank vent lines and the vent header must, of course, accommodate the higher throughput. This is easily accom-modated by an increase in pipe size, e.g. an increase from 2" to 3" pipe will allow an increase from 16 CFM at a frictional line loss of 0.25 inches of water per 100 feet to 35 CFM at the same line loss. The impact on the design of the organic control device is however substantial.

The control device must be designed for a flow rate that is nearly three times that required to handle breathing losses. Such control devices as carbon canisters and vapor condensers have maximum allowable gas velocities or more accurately must exhibit a certain ratio of condensing or adsorbing surface area per cubic feet of laden gas. If the throughput triples the device's capacity must triple. Such devices as boilers or incinerators have similar limitations though normally these devices are present for another purpose and usually have sufficient extra capacity.

Plumbing the System

First it must be decided what storage tanks can be vented into common headers. The concern here is that incompatible vapors may create problems. In a hazardous waste facility that distills solvents for reuse, co-mingling non-chlorinated vapor and chlorinated vapors may cause product quality problems in the recovered solvents. If certain liquids are kept separated it is advisable to keep their vapors separated.

Each storage vessel will have a vacuum relief vent, a normal relief vent, (normally set to relieve at 6 or 8 inches of water pressure) and an emergency escapement vent (normally set to relieve at 10 to 12 inches of water pressure). It is best if the closed vent system vent line is connected at a separate tank top nozzle. This system vent line must have a high point elevation higher than the normal vent or the escapement vent. If the storage tanks are "inerted", that is, only an inert gas such as nitrogen can be drawn into the tank, a flame arrestor is not needed on this system vent line. If a flame arrestor is installed it needs to be accessible for inspection. The size of the vent line should be no smaller than the largest liquid transfer line entering the tank. The tank vent line should enter into the top of the vent header and the header should be sloped to a low point liquid drain. This reduces the possibility of having liquid present to act as a "liquid seal" or otherwise obstruct the system.

The most frequently encountered problem is to end up with a closed-vent system that vents out through one or more of the tank vents rather than through the vapor control device. This may simply be the result of defective relief vents on tanks. However it may be more complex. The pressure drop between the tank and the exit of the vapor control device must be less than the tank relief vent setting. This pressure drop is made up of three components.

  1. The line loss between the tank and the header. Unless this line is especially long or restricted this is not usually the problem.

  2. The line loss in the header to the control device. This is frequently the problem due to greater than design gas flows, most often when performing two or three simultaneous transfers on a hot day. To prevent this from happening consideration must be given to designing this header for maximum expected gas flow.

  3. Insufficient capacity of the vapor control device, at the operating conditions of the system. Vapor control devices are rated by throughput of gas at specified conditions with a specified concentration of vapors. At higher throughput rates the pressure differential across the device will rise. If the pressure rise is great enough one or more tank relief vents may relieve. To prevent this from happening either:

    1. The vapor control device must be sized based on the worst case conditions (i.e. maximum expected gas flow rate at highest temperature and lowest system pressure) or

    2. The vapor control device must be designed to handle the maximum expected gas flow rate at the highest temperature and at a specific pressure. The vent system must then be designed to deliver the gas at that pressure. Pressure is very important. The higher the pressure at which the system can operate, the more efficiently the system will operate. I have seen systems operate at nearly one psig, about two feet of water pressure. The difficulty with this is that this is the upper limit of allowable internal pressure for most atmospheric tanks and it is very difficult to find relief vents rated at this pressure.

Most of the closed vent systems that are currently installed are constructed of normal weight pipe, i.e. rated for 150 psig. This is not necessary however, pneumatic steel tubing rated for full vacuum and 15 psig will be more than sufficient. Such tubing is available in plain steel, galvanized or even stainless steel. Regardless of the material of construction the pipe is usually threaded or welded together. Welded connections, however, reduce Subpart BB monitoring requirements. There are very few valves in the systems, usually only those needed to valve-off the transport container vapor connections. It is recommended that the connection to the transport container be open ended and capped when not in use or alternately use one of the special connections that incorporates a spring loaded valve that is forced open when connection is made. This precludes a valve being inadvertently left closed. If there is a need to isolate some part of the system on a routine basis the use of flange blinds or lockable valves is a preferred method.

Selection of the Vapor Vent Control Device

For small systems with transport container vapor vent back capabilities the two drum carbon canister systems predominate. The carbon canisters are 55 gallon drums fitted with entry and outlet connections, a gas distribution grid and one to two feet of activated carbon. The vapor laden gas enters near the bottom of the drum travels up through the bed of carbon and exits out the top of the drum. A second drum is placed in series with a sample port between them. Be aware there is a considerable range in "loading" capacities for the various grades and mesh sizes of activated carbon. Also as materials are absorbed onto activated carbon a certain amount of heat will be generated. The higher the loading the more heat is generated. Some materials especially ketones (e.g. methyl ethyl ketone, methyl isobutyl ketone) have been known to set activated carbon canisters on fire. Some grades and mesh sizes are less of a problem. Also, carbon that has been in service for a while, particularly in humid regions, is less likely to catch on fire. Additional measures such as carbon bed thermometers or high carbon bed temperature alarms are a worth while consideration for facilities that routinely handle materials that contain ketones. Also manual or automated injection of water into the carbon bed will add significant fire protection. The monitoring and reporting requirements for non-regenerated carbon canisters are relatively simple.

Those facilities that have systems that vent to regenerated activated carbon systems, refrigerated condenser systems or are ducted into incinerators or BIFs, generally have process vents (i.e. distillation systems, blending/grinding system vents, drum waste removal system vents) that require such high capacity systems. Similarly for vents that are ducted to incinerators, or BIFs. The monitoring and reporting requirements for these vapor control devices are much more complex than for disposable carbon systems.

As stated earlier, designing and constructing one of these systems to accomplish control of the vapors is not especially difficult. The difficulty is encountered in designing systems that minimize maintenance costs and ease compliance with the numerous state and federal regulations. Before beginning this task the designer should have a thorough understanding of the regulations.