Kevin S. Fisher, Curtis Rueter
Radian Corp.
Austin, Tex.
Mel Lyon
Public Service Co. of
Colorado Denver
Jorge Gamez
Gas Research Institute
Chicago
For natural gas glycol dehydrators, evaporative cooling provides better emission control than conventional air cooled or glycol-cooled condensation technologies.
Recent developments in environmental regulations, including the upcoming Maximum Achievable Control Technology (MACT) standards, have made emissions of benzene, ethylbenzene, toluene, and xylenes (BTEX) from glycol dehydrators a major concern for the natural gas industry.
This concern has led the gas industry to look for new ways to reduce and control hydrocarbon emissions from these systems. Gas Research Institute (GRI) has sponsored the development of the R-BTEX process, which uses evaporative cooling.1
EMISSION CAUSES
A typical glycol dehydration unit (Fig. 1) consists of an absorber, a flash tank, heat exchangers), a glycol reboiler still, and associated pumping and piping equipment.
In the absorption step of the dehydration process, BTEX and other volatile organic compounds (VOC) are also absorbed from the natural gas. Triethylene glycol (TEG), the most common glycol solvent, has a particularly high affinity for BTEX compounds.
From the absorber, the glycol flows to the flash tank, or separator, which is typically present only on larger units. The flash tank or separator removes a large fraction of the light gases by reducing the pressure. The flash gas may be used as fuel.
Because flash separators typically operate at 100-160 F. and 30-90 psig, only a small fraction (usually less than 5%) of the BTEX in the glycol is removed.2 The normal boiling points for BTEX compounds range from 176 to 284 F.
Following the flash separator, the glycol is distilled in a regenerator to strip water that consequently removes BTEX and other VOC as well. The still vent stream is typically 50 wt % water, with the balance being hydrocarbons along with small quantities of carbon dioxide, nitrogen, and hydrogen sulfide, if present in the processed gas.
The still vent stream is usually released to the atmosphere, while the recovered lean (dry) glycol is recycled for use in the absorber.
EMISSION CONTROL
Emission control technologies have become of greater interest to the industry as a result of increasing regulatory pressure.3 The Clean Air Act Amendments (CAAA) of 1990 have been the impetus for air toxics regulations, and the maximum achievable control technology (MACT) standards for the oil and gas industry are expected to be proposed in June 1995 and promulgated in June 1996. Recent surveys estimate that there are over 40,000 glycol dehydrator units in the 13.s.'
BTEX compounds have been listed as hazardous air pollutants (HAP) in the 1990 CAAA. According to the CAAA Section 112(a) definition, sources emitting over 10 tons/year of any single HAP or over 25 tons/year of any combination of HAP are defined as major sources subject to regulatory emission standard setting activities.
Many glycol dehydration units may be classified as major sources based on this definition. Sources emitting less than the 10-25 tons/year limits are defined as area sources. EPA has estimated that most of the HAP emissions from dehydrators may arise from area sources. Because of the large number and amount of emissions from smaller dehydrators, EPA may also regulate glycol dehydrators as area sources and may apply MACT standards to these units.'
In addition to these federal regulatory developments, several states are already regulating emissions from glycol dehydrators. For example, Louisiana has three rules: Waste Gas Disposal Rule (LAC:III.2115), Comprehensive Toxic Air Pollutant Emission Control Program (LAC:33.III.Chapter 51), and a Specific Glycol Dehydrator Rule (LAC:33.III.2116).
The Waste Gas Disposal Rule and the comprehensive toxic air regulations apply to relatively large sources and require high emission control efficiencies. The glycol dehydrator rule applies to units emitting greater than 9 tons/year VOC and requires less stringent control.
Ventura County, Calif., has recently approved a rule requiring 95% control of reactive organic compound (ROC) emissions from glycol dehydrators. ROC include all hydrocarbons except methane.
New Mexico, Colorado, and Oklahoma have no specific regulations concerning glycol dehydrators but are requiring dehydrators to be permitted under general air toxics regulations. Texas and Wyoming are also considering specific regulations for emissions from glycol dehydrators.
It may be possible to reduce emissions to acceptable levels at some units by optimizing dehydrator operations. Reducing the glycol circulation rate is one simple way to lower emissions. There is a linear relationship between the glycol circulation rate and BTEX emissions.
Gas temperature and BTEX content are two other factors that significantly influence the amount of BTEX absorbed into the glycol, and consequently the level of emissions.
AVAILABLE CONTROLS
Because some dehydrators may produce emissions that are above regulatory limits even at optimum operating conditions, several technologies have been developed to control emissions from glycol dehydrators. The two most common methods of controlling emissions are condensation and combustion (or incineration).
Vent-stream combustion is difficult, however, because of the possibility of hydrocarbon and water condensing in the flare. Condensation in the flare may result in smoking and incomplete vapor combustion.
Operating costs for flares and incinerators are relatively high because no hydrocarbons are recovered for sale and supplemental fuel is often required to sustain an adequate flame temperature.
For these reasons, condensation has become the most common method for controlling glycol-dehydrator emissions.
AIR-COOLED CONDENSERS
Air-cooled condensers with either natural or forced-draft cooling are one of the most common options because of the relatively simple design and competitive capital cost. However, the thermodynamic limitations require the condenser operating temperature to be higher than the ambient dry-bulb air temperature.
For most practical heat-exchanger designs, the condenser temperature is about 20 F. higher than the ambient air.
GLYCOL-COOLED CONDENSERS
Glycol-cooled condensers use rich glycol from the dehydrator, prior to any three-phase separator, as the coolant. The condenser outlet temperature is dependent on the rich-glycol temperature, which is in turn dependent on the natural gas temperature that is typically at or above ambient temperature.
Normal approach temperatures for liquid-cooled exchangers may be about 10-20 F. Because the rich glycol may already be above ambient temperature, it is likely that the outlet condenser temperature will be at least 10-25 F. above ambient. Glycol-cooled condenser temperatures may be much higher if adequate heat-exchanger surface area is not provided.
WATER-QUENCH CONDENSERS
Water-quench condensers combine a cool water stream with the dehydrator still-vent stream to condense the water and hydrocarbons. The quench water is cooled in a separate heat exchanger, most likely air cooled, or refrigeration system.
A refrigeration system achieves lower condenser temperatures but requires additional electricity and mechanical complexity.
EMISSION RATES
Emission rates increase rapidly with condenser operating temperature. The condenser temperature will dictate the emission control performance for a given condenser and regenerator vent stream. Because the vent stream contains methane and other noncondensable gases, some of the VOC will remain uncondensed in the gas phase.
At higher temperatures, the VOC have a higher vapor pressure resulting in a larger fraction of these compounds remaining uncondensed in the gas phase.
Fig. 2 shows how emission rates will increase with condenser temperature for a typical glycol regenerator vent stream. These estimates were made using GRI Glycalc, a PC-based program for estimating emissions from glycol dehydration units.
The base case operating conditions and gas composition data are shown in Table 1. For each of the components or categories shown, the amount of vapor not condensed is about doubled with each 20 F. increase in temperature.
R-BTEX PROCESS
All of the previously described condensation technologies may meet regulatory requirements depending on site specific operating conditions. However, these technologies are limited to above ambient operating temperatures unless a separate refrigeration system is supplied. The evaporative cooling of the R-BTEX process achieves subambient condenser temperatures and better emission control without needing separate refrigeration equipment.
Fig. 3 shows a flow diagram for the R-BTEX process. The dehydrator still-vent vapors first pass through an air-cooled heat exchanger in which most of the water and some of the hydrocarbons are condensed. The partially condensed stream then goes to a water-cooled heat exchanger where most of the remaining hydrocarbon vapors condense.
The condensed, three-phase stream flows to a separator, where it is partitioned into noncondensable gas, hydrocarbon liquids, and water. The small amount of noncondensable gas can be:
- Vented directly to the atmosphere if the amounts of BTEX and other VOC contained in the vent gas are below regulatory limits.
- Used as fuel gas in the reboiler or sent to a flare.
The hydrocarbon layer is decanted and stored for sale. The condensed water is returned to the cooling system as make-up water, and excess water is blown down as needed to maintain the water balance.
Utility requirements for the R-BTEX process are minimal. Water may be supplied at process start-up or can be recovered initially by running only the air-cooled exchanger. A small amount of electricity is needed for the cooling tower fan and for the cooling pump.
The R-BTEX process's key feature, which distinguishes it from other condensation processes, is that cooling water is generated within the process by recovering steam from the glycol dehydrator. The water is cooled to below ambient temperature (dry-bulb temperature) by evaporating water across the cooling-tower packing.
The cooling-water temperature will typically approach within a few degrees of the wet-bulb temperature. The wet-bulb temperature is below ambient temperature except at 100% relative humidity.
The wet-bulb temperature is the dynamic equilibrium temperature attained by a water surface when the rate of heat transfer to the surface is balanced by the rate of mass transfer away from the surface,'
Under most conditions the wet-bulb temperature is equivalent to the adiabatic (constant enthalpy) saturation temperature. By contrast, the dew point occurs at the point where the air is saturated without a change in water content.
Cooling water, at or near wet-bulb temperature and less than ambient, allows maximum recovery of hydrocarbons because the final condenser temperature is typically below ambient. This results in higher levels of control than can be achieved with processes that operate at or above ambient temperatures.
HYDROCARBON RECOVERY
Hydrocarbon recovery can favor condensation processes over incineration at sites where hydrocarbons can be collected and sold for profit.
Fig. 4 shows that the value of recoverable hydrocarbons increases with the dehydrator throughput and with increasing BTEX content in the gas. This graph was generated using GRI Glycalc program for the dehydrator operating conditions and gas composition data shown in Table 1.
CASE STUDY
Public Service Co. of Colorado installed an R-BTEX unit to control emissions from a TEG dehydrator at its Baxter Pass compressor station (Fig. 5). The plant processes 5-10 Mscfd of sweet gas produced from the Dakota formation in western Colorado. The station provides compression, dehydration, and recovery of C6, compounds for hydrocarbon dew point control. Most of the gas processed at the station is distributed for local residential use.
In 1993, Public Service Co. was considering options for controlling dehydrator emissions at the Baxter Pass station and agreed to support GRI's field evaluation program to collect performance and reliability data for the R-BTEX process at several commercial installations.
The initial test, conducted June 1994, characterized emission control performance. Subsequent testing, through November 1994, evaluated the reliability of the unit and monitored emission control performance over time.
Because of the harsh winter conditions at this site, the R-BTEX unit was housed in a small metal building. An auxiliary gas-fired heater also was provided to prevent the condensed water from freezing during extreme periods of cold weather.
Table 2 shows the dehydrator process conditions at the site during the June 1994 test. Wet gas from the compressors entered the absorber at about 800 psig and 100 F. Glycol was circulated with a gas-operated pump at 172 gal/hr. The unit was also equipped with a flash tank that recovered methane and other light ends for regenerator fuel gas.
The composition of the inlet gas to the dehydrator is shown in Table 3. Table 4 gives the regenerator overheads composition and flow rate, based on measurements made at the site during the testing. This table shows that water by weight makes up only about 37% of the stream. More than half of the stream is composed of condensable hydrocarbons.
Table 4 also shows that BTEX components in the regenerator vent stream occur in greater proportion than in the natural gas (Table 2). This result is typical of TEG dehydrators and occurs because of TEG's affinity for aromatic compounds.
Table 5 shows the emission control performance for the R-BTEX unit during the performance testing. The uncontrolled emission rate for VOC- was 166 tons/year. The controlled VOC emission rate was less than 10 tons/year, corresponding to a 94% VOC control efficiency.
BTEX emissions were controlled from 62 to less than 2 tons/year, with a corresponding 97% control efficiency. The lower control efficiency for VOC compounds reflects noncondensable VOC compounds, such as propane, that are present in the regenerator overheads stream.
The relatively high control efficiencies shown in Table 5 result from subambient condenser temperatures provided by the R BTEX evaporative cooling system. Fig. 6 shows a comparison of the condenser outlet temperature and the ambient temperature at the site during the test period from March to December 1994.
The condenser outlet temperature was significantly lower than the ambient temperature at the site. These subambient process temperatures are a key to obtaining maximum emission control performance without needing auxiliary refrigeration systems. On average, the condenser temperature was 11 F. below ambient temperature during the test period.
The condenser temperature became increasingly subambient at the higher ambient temperatures. This is an important performance characteristic because at low ambient temperatures the subambient condenser temperatures are not as critical in achieving a high level of control. In contrast, the condenser temperature becomes important during hot weather when it is necessary to operate at as low a temperature as possible.
Fortunately, hot weather is frequently accompanied by low humidity at many locations. During the hottest ambient conditions of near 100 F., the condenser was operating around 80 F. A comparable air cooled or glycol-cooled condenser might operate at 110-120 F. under these conditions, resulting in substantially higher emission rates. These performance characteristics show that the R-BTEX process has an even stronger advantage over other processes in locations with hot, dry climates.
With these low condenser temperatures, about 3.2 b/d of salable hydrocarbon liquids were recovered. The corresponding projected annual revenue is about $20,000, based on a typical market price of $17/bbl.
Operating expenses were minimal, consisting of a few hours a week to monitor the unit and about 1 kw of electricity.
The payback period for the unit is estimated at 2-3 years, based on a purchased equipment price of $42,000 which included the heated building and fire-tube water heater needed for winterizing the facility.
REFERENCES
1. Sivalls, C.R., Rueter, C.O., Fisher, K.S., and Gamez, I.P., "The R-BTEX Process for Mitigating Air Emissions from Glycol Dehydrators," Paper No. GRI 94/0156, GRI Glycol Dehydrator/Gas Processing Air Toxics Conference, Austin, Tex., April 1994.
2. Thompson, P.A., Cunningham, I.A., Berry, C.A., and Evans, J.M., "PC program estimates BTEX, VOC emissions," OGJ, June 14, 1993, pp. 36-41.
3. Rueter, C.O., Wessels, J.K., and Hurtado, M.L., Paper No. GRI 94/0156, GRI Glycol Dehydrator/Gas Processing Air Toxics Conference, June 1994.
4. Graham, J.F., Krenek, M.R., Maxson, D.J., Pierson, J.A., and Thompson, J.L., Final Report No. GRI 94/0099, "Natural Gas Dehydration: Status and Trends," January 1994.
5. Fitzsimons, G., and Viconovic, G. "Status of U.S. Environmental Protection Agency Activities on Oil and Gas Production MACT Standard Development," GRI Glycol Dehydrator/Gas Processing Air Toxics Conference, Austin, Texas, April 1994.
6. Perry, R.H., and Chilton, C.H., editors, Chemical Engineer's Handbook, Fifth Edition, McGraw-Hill, 1973, P. 12-2.
Copyright 1995 Oil & Gas Journal. All Rights Reserved.
