Accurate measurement and prediction of crude oil and natural gas liquid (NGL) products vapor pressure are important for safe storage and transportation, custody transfer, minimizing vaporization losses and environmental protection. Vapor pressure specifications are typically stated in Reid Vapor Pressure (RVP) or/and True Vapor Pressure (TVP). In addition to the standard procedures for their measurements, there are rigorous and shortcut methods for their estimation and conversion.

There are figures and monographs for conversion of RVP to TVP for NGLs (Natural Gas Liquids) and crude oil at a specified temperature. This tip will present simple correlations for conversion from RVP to TVP and vice versa at a specified temperature. The correlations are easy to use for hand or spreadsheet calculations. Figures generated using these correlations will be presented, too.

Chapter 5 of reference [1] presents an excellent overview of TVP and RVP including their definitions, standard procedures for their measurements and diagrams for their conversion. The proceeding two paragraphs are extracted with minor revisions from reference [1].

TVP is the actual vapor pressure of a liquid product at a specified temperature and is measured with a sample cylinder. TVP specifications must always be referenced to a temperature, which frequently falls between 30-50°C (86-122°F). TVP is difficult to measure and depends on the ratio of the vapor to liquid, V/L, in the measurement device. If V/L = 0, the vapor pressure is essentially equivalent to the bubble point of the mixture which is the highest vapor pressure value for the liquid. As V/L increases, i.e., a small amount of vapor exists at the point of measurement, the measured vapor pressure will decrease. The relationship between the measured TVP and V/L depends on the composition of the mixture. For “near pure” component mixtures, V/L has little effect on the measured vapor pressure. For mixtures with a large composition range, such as crude oil or condensate, the effect of V/L on the measured vapor pressure can be significant (See ASTM D 6377 – 10 for detail). Reference [1] lists the Standards used for TVP measurements.

Because of the difficulty in accurately measuring TVP, an alternative method of measuring vapor pressure is frequently used. This is the RVP.  The RVP is a standard test set out in ISO 3007:1999, Petroleum products and crude petroleum – Determination of vapour pressure – Reid method. Another standard that applies to RVP is:  ASTM D323 – 08, Standard Test Method for Vapor Pressure of Petroleum Products (Reid Method). The RVP test is applicable for crude oils, condensates, and petroleum products such as gasoline (petrol) mixtures. A liquid sample is collected in the lower 20% chamber (see Fig 5.12 of reference [1]). The 80% chamber, which is filled with air (may also contain a small amount of water vapor) at atmospheric pressure.  Both chambers are cooled to 0°C (32°F). The 80% air chamber (at atmospheric pressure) is then connected to the 20% liquid chamber. The connecting valve is opened and the cylinder is heated in a water bath to 37.8°C (100°F). The pressure indicated on the gauge is the RVP.

Reference [1] provides Figure 5.14 for conversion RVP to TVP for motor gasoline (petrol) and natural gasoline (C₅₊ NGLs) at various temperatures. Figure 5.15 of reference [1] is a nomograph that shows the approximate relationship between RVP and TVP for crude oil. It is commonly used for converting RVP to TVP at custody transfer points where the vapor pressure specification for the oil is a TVP, but the actual vapor pressure measurement is an RVP.

Vazquez-Esparragoza et al. [2] present an algorithm to calculate RVP without performing the actual test. The algorithm, based on an air-and-water free model, uses the Gas Processors Association Soave-Redlich-Kwong [3] equation of state and assumes liquid and gas volumes are additive. This algorithm can be used to predict RVP of any hydrocarbon mixture of known composition. Since the calculations are iterative, it should be incorporated into a general purpose process simulator. Vazquez-Esparragoza et al. [2] reported good agreement between predicted and experimental values.

Riazi et al. [4] presented a new correlation for predicting the RVP of gasoline and naphtha based on a TVP correlation. The input parameters for this correlation are the mid-boiling point, specific gravity, critical temperature, and critical pressure, where the critical properties may be estimated from the boiling point and specific gravity using available methods. They evaluated their proposed correlation with data collected on 50 gasoline samples from crude oils from around the world with API gravity ranges from 41 to 87, average boiling point ranges from 43 to 221 °C (110 to 430 °F) and RVP of 0.7 to 115 kPa (0.1–17 psi). The average error from their proposed correlation is about 6 kPa (0.88 psi) [4].

Development of Model for Motor Gasoline and Natural Gasoline

To develop the desired correlations for conversion of motor gasoline and natural gasoline RVP to TVP and vice versa, this tip generated 127 data points from Figure 5.14 of reference [1]. These data covered the full ranges of temperature, RVP and TVP of Figure 5.14.

RVP to TVP: This tip used these data to determine the parameters to the Equations 1 through 3. The API 2517 originally reported the same forms of equations for crude oil.

TVP to RVP: Similarly this tip proposes the following equations for conversion from TVP to RVP.

where:

T          = Temperature, °C (°F)

RVP    = Reid Vapor Pressure, kPa (psi)

TVP    = True Vapor Pressure, kPaa (psia)

Note that the values of A₁, A₂, B₁, and B₂ are different in the above two sets of equations. The value of “C” is a function of the chosen units (SI versus FPS) and is consistent.

Table 1 presents the optimized values of A₁, A₂, B₁, B₂, and C for three sets of data in FPS (Foot-Pound-Second) and SI (System International). The data set labeled “All” included 127 data points covering all of the data of combined motor gasoline and natural gasoline. The data set “Motor Gasoline’ and “Natural Gasoline’ cover 76 and 51 data points for motor gasoline and natural gasoline, respectively. Table 1 also presents the Average Absolute Percent Deviation (AAPD), the Maximum Absolute Percent Deviation (MAPD), Average Absolute Deviation (AAD), and the number of data points (NP) for each data set. The error analysis of Table 1 indicates that the accuracy of the proposed correlations is good. Their accuracy is as good as the quality of tabular data generated from Figure 5.14.

Table 1. The optimized parameters for motor gasoline and natural gasoline

¹AAPD           = Average Absolute Percent Deviation
²MAPD           = Maximum Absolute Percent Deviation
³AAD              = Average Absolute Deviation
⁴NP                 = Number of data Points (NP)

Figures 1a (SI) and 1b (FPS) present the tabular data (generated from Figure 5.14 of reference [1]) in the form of “legends”. This figure also presents the predicted TVP by the proposed correlations (Equations 1 through 3 and the corresponding parameters listed in Table 1) as a function of temperature and RVP in the form of “continuous lines” and “broken lines” for motor gasoline and natural gasoline, respectively.

Development of Model for Crude Oil

To develop the desired correlations for conversion of crude oil RVP to TVP and vice versa, this tip generated 196 data points from Figure 5.15 of reference [1] using the correlations reported in API 2517. The API 2517 equations are the same as shown in Equations 1 through 3. These data covered the full ranges of temperature, RVP and TVP of Figure 5.15.

RVP to TVP: Table 2 presents the API 2517 FPS parameters of A₁, A₂, B₁, B₂, and C for Equations 1a, 2a, and 3a (T, °F, RVP, psi and TVP, psia). This tip determined the SI (T, °C, RVP, kPa and TVP, kPaa) corresponding parameters. Table 2 also presents the SI parameters.

Table 2. API 2517 parameters for crude oil TVP calculations
(For use with Equations 1a, 2a and 3a)

Figures 2a (SI) and 2b (FPS) present the predicted TVP by Equations 1a, 2a and 3a and the corresponding parameters listed in Table 2 as a function of temperature and RVP for crude oil.

TVP to RVP: To cover the full ranges of Figure 5.15 of reference [1], this tip proposes the following equations for conversion of TVP to RVP.

where:

T          = Temperature, °C (°F)

RVP    = Reid Vapor Pressure, kPa (psi)

TVP    = True Vapor Pressure, kPaa (psia)

Table 3 presents the optimized values of A₁, A₂, A₃, B₁, B₂, B₃, and C for RVP calculation in FPS and SI for the proposed Equations 1c through 3c. Table 3 also presents the Average Absolute Percent Deviation (AAPD), the Maximum Absolute Percent Deviation (MAPD), Average Absolute Deviation (AAD), and the Number of data Points (NP). The error analysis of Table 3 indicates that the accuracy of the proposed correlations is good and can be used for the estimation purposes.

Table 3. The optimized parameters for crude oil TVP conversion to RVP

(For use with Equations 1c, 2c & 3c)

1AAPD           = Average Absolute Percent Deviation

2MAPD           = Maximum Absolute Percent Deviation

3AAD              = Average Absolute Deviation

4NP                 = Number of data Points (NP)

 

Conclusions:

For converting RVP to TVP of motor gasoline and natural gasoline, this tip presented simple correlations similar to the ones reported in API 2517 for crude oil RVP to TVP. This tip determined the correlation parameters by regressing the data generated from available diagrams. These correlations were also extended for easy conversion from TVP to RVP. Tables 1, 2, and 3 present the correlation parameters in SI and FPS system of units. Tables 1 and 3 presents the accuracy of the proposed correlations against the data generated from diagrams. Tables 1 and 3 indicate that the accuracy of these correlations is as good as the quality of the data in the original diagram and can be used for easy conversion of RVP to TVP or vice versa. These correlations are easy to use for hand or spreadsheet calculations and should be used for estimation purposes. For accurate measurements, correction procedures outlined in ASTM D 6377–10 and other guidelines should be consulted.  Several organizations are currently working to improve the accuracy of TVP estimation from RVP and/or VPCR(x) (ASTM D6377) measurement techniques. In all cases, Federal and State Laws and Regulations should be followed for safety and environmental protection.

To learn more about similar cases and how to minimize operational problems, we suggest attending our G4 (Gas Conditioning and Processing), G5 (Advanced Applications in Gas Processing), and PF4 (Oil Production and Processing Facilities), courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

By: Dr. Mahmood Moshfeghian

References:

1. Campbell, J.M., Gas Conditioning and Processing, Volume 1: The Basic Principles, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.
2. Vazquez-Esparragoza, J.J., et al, “How to Estimate Reid Vapor Pressure (RVP) of Blends”, Bryan Research and Engineering, Inc, Bryan, Texas, 2015.

3. Soave, Chem. Eng. Sci. 27, 1197-1203, 1972.
4. Riazi, M.R., Albahri, T.A. and Alqattan, A.H., “Prediction of Reid Vapor Pressure of Petroleum Fuels”, Petroleum Science and Technology, 23: 75–86, 2005.

The December 2012 Tip of the Month (TOTM) [1] discussed the hydrate phase behavior of sour natural gas mixtures. Specifically, it showed carbon dioxide inhibits the hydrate formation slightly while hydrogen sulfide enhances hydrate formation considerably. This tip will extend the previous study on the natural gas hydrate formation phase behavior. Specifically, it will study the impact of nitrogen on the formation of hydrate in a natural gas mixture.

The hydrate formation temperature of a gas depends on the system pressure and composition. There are several methods of calculating the hydrate formation conditions of natural gases [2-5]. References [2-3] present rigorous methods while [4-5] present the shortcut methods suitable for hand calculations. This study uses a rigorous method using the Soave-Redlich-Kwong (SRK) equation of state [6] in ProMax [7] software.

Table 1 presents the compositions of the gas mixture studied. Notice that in each case about 20 mole % of methane is replaced with the same amount of either nitrogen, carbon dioxide or hydrogen sulfide.

Table 1. Water-saturated compositions of gas mixtures studied

Figure 1 presents the calculated hydrate formation curve (broken curve) and the dew point portion of the phase envelope of a sweet natural gas (continuous curve). Figure 1 also presents the dew point and hydrate formation curves for the gas mixture containing 20 mole % nitrogen (N₂).

Figure 1 indicates that the presences of 20 mole % N₂ shifts the hydrate formation curves slightly to the left, depressing the hydrate formation temperature. Note that the points to the left and above the hydrate curves represent the hydrate formation region. From an operational point of view, this region should be avoided. This figure also indicates that the presence of N₂ increases the cricondenbar and the two-phase (gas + liquid) region within the envelope expands.

Figure 2 presents the calculated hydrate formation curves for a sweet gas (Continuous curve) with no N₂, sour gases containing 20 mole % CO₂ or H₂S, and a sweet gas containing 20 mole % N₂ (broken curves). This figure clearly indicates that the impact of N₂ is much less than of H₂S and slightly less than of CO₂. Nitrogen and carbon dioxide depresses the hydrate formation condition slightly (shift the hydrate curves to the left) but H₂S promotes hydrate formation considerably. As an example, at 1000 psia (6900 kPa), N₂ reduces hydrate formation temperature for this sweet gas by about 4.5˚F (2.5˚C), CO₂ reduces hydrate formation temperature by about 5.5˚F (3˚C) while, H₂S increase the hydrate formation temperature by about 20˚F (11.1˚C).

Conclusions:

Katz and co-workers [8] developed a set of vapor-solid equilibrium constants (K_v-s) values for hydrate prediction. In the Katz method as described on page 161 of Chapter 6 of reference [6] “nitrogen is a hydrate former, and it is likely that some nitrogen may end up in the hydrate lattice in typical natural gas production systems. However, it is not a factor in determining hydrate formation conditions unless you are working with mixtures of nitrogen and methane which are sometimes found in coalbed methane production. In these cases the N₂-CH₄ mixture will have a lower hydrate formation temperature than pure methane. As a practical matter using K_v-s = (infinity) for nitrogen gives satisfactory results for typical natural gas mixtures”.

This study has showed that the presence of N₂ and CO₂ and H₂S in natural gas has an opposite impact on the hydrate formation condition. While the impact of N₂ and CO₂ is small in the same direction, H₂S has considerable impact on the hydrate formation condition in the opposite direction. For the same composition and condition studied, nitrogen and carbon dioxide slightly depresses hydrate formation (acts as hydrate inhibitor and shifts the hydrate curve to the left) while H₂S shifts the hydrate curve to the right considerably, promoting hydrate formation conditions, and may cause severe operational problems.

To learn more about similar cases and how to minimize operational problems, we suggest attending our G4 (Gas Conditioning and Processing), G5 (Advanced Applications in Gas Processing), P81 (CO₂ Surface Facilities), and PF4 (Oil Production and Processing Facilities), courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

By: Dr. Mahmood Moshfeghian

Reference:

1. Moshfeghian, M. http://www.jmcampbell.com/tip-of-the-month/2012/12/sour-gas-hydrate-formation-phase-behavior/
2. Parrish, W.R., and J.M. Prausnitz, “Dissociation pressures of gas hydrates formed by gas mixtures,” Ind. Eng. Chem. Proc. Dev. 11: 26, 1972.
3. Holder, G. D., Gorbin, G. and Papadopoulo, K.D, “Thermodynamic and molecular properties of gas hydrates from mixtures containing methane. argon, and krypton,”  Ind. Eng. Chem. Fund. 19(3): 282, 1980.
4. Campbell, J.M., Gas Conditioning and Processing, Volume 1: The Basic Principles, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.
5. Gas Processors Suppliers Association; “ENGINEERING DATA BOOK” 13^th Edition – FPS; Tulsa, Oklahoma, USA, 2012.

6. Soave, Chem. Eng. Sci. 27, 1197-1203, 1972.
7. ProMax 3.2, Bryan Research and Engineering, Inc, Bryan, Texas, 2015.
8. Carson, D. B. and D. L. Katz, Trans. AIME, Vol. 146, p. 150, 1942.

In the post-World war II period, the steels used in the oil and gas industry were quite different from what we use today. This tip of the month (TOTM) presents a brief overview of improvements in the steels used in oil and gas processing equipment for safer and more reliable operations.

Plate was SA-285C a 55,000 psi (379 MPa) tensile steel that was relatively soft and easy to fabricate. It was not killed steel and therefore, not fine grain steel. The low tensile strength meant thicker vessels and because of poor welding techniques, spot or no radiography at all was common, making the items even thicker. Figure 1 shows a vacuum tower made of SA-285C from the 1950’s. This tower was constructed in 1961 by Chicago Bridge and Iron for the Shell Martinez Refinery in California.

A plate designated SA-212B Firebox was in use for higher tensile applications. It had a 70,000 psi (482 MPa) tensile, but was coarse grained and had the undesirable characteristic of fracturing in the parent metal after thermal expansion and contraction over a period of time. Due to repeated failures in service, this material was removed from the ASME Boiler and Pressure Vessel Code Section II in 1968 as being unfit for thermal cycling.  Figure 2 presents a high pressure molecular sieve tower which was fractured by thermal cycling.

Pipe used in the 1950’s was SA-53B, which could be Electric Resistance Welded or seamless. It was not killed steel. It had a 60,000 psi (413 MPa) tensile and was the pipe of choice for vessel, tank, and piping fabrication at the time.

The forging of the 1950’s was SA-181, a 60,000 psi (413 MPa) tensile steel used for flanges, forged steel fittings, and heavy nozzles. It was not killed steel.

Since none of these steels were killed, fine grained steels, their use declined rapidly as the industry moved into harsh environments such as the North Slope of Alaska and the processing of acid gases and sour crudes.

Killed steel came into wide use during the 1960’s. Killed steel is produced in the ladle by adding silicon or aluminum to prevent further deoxidation of the heat. Molten steel contains dissolved oxygen which can cause bubbles in the cooling and solidification process. The addition of silicon or aluminum stops the reaction of the oxygen with carbon, producing a fine grain steel free from dissolved gases, highly homogenous with excellent fabrication properties.

During the 1960’s, the SA-516 family of plate steels was introduced. These steels were silicon killed, fine grained, and produced excellent properties. The fine grain gave the steel impact resistance at temperatures down to -50 °F (-45.5 °C). The SA-516 suffix defines the tensile strength, 55,000, 60,000, 65,000, and 70,000 psi (379, 413, 448, and 482 MPa).

• SA-516-55 was designed to replace SA-285C
• SA-516-60 was designed for use in very cold service.
• SA-516-65 was for intermediate tensile requirements
• SA-516-70 was to replace SA-212B Firebox plate

The chemical and mechanical properties of these four grades of steel overlap to the extent that one plate can actually meet all four specifications.

Approximately 90% of all custom carbon steel pressure vessels manufactured for the oil and gas industry in the world today are made from SA-516-70 or its UNS (Unified Numbering System) equivalent. Figure 3 presents an example of a vertical drum made of SA-516-70.

During the 1960’s SA-106 pipe replaced SA-53 as the pipe of choice.  Unlike SA-53B, SA-106B is seamless, killed, fine grain steel. It has a 60,000 psi (413 MPa) psi tensile.

In 1978, SA-105 forgings replaced the SA-181 as the forging material of choice. SA-105 has a tensile of 70,000 psi (482 MPa), so the pressure ratings of B16.5 carbon steel flanges increased.

Around the year 2000, the pipe manufactures improved their processes making SA-106 pipe to the point that they are able to meet the chemical and mechanical properties of SA-106B and SA-106C in the same heat.

Since 2003, basically all SA-106 pipe is dual certified to SA-106B and SA-106C. This means that all three major components of a pressure vessel or shell and tube heat exchanger now have the same tensile strength, 70,000 psi (482 MPa). Figure 4 presents pipes made of SA-106.

Austenitic Stainless steels (300 series) fifty years ago were made to straight grade (0.08 carbon) or “L” grade (0.03 carbon). Steel service centers had to maintain stocks of both grades. About 45 years ago, the stainless mills improved their manufacturing techniques to produce dual certified stainless steel, meaning that virtually all stainless in the steel service centers meets the criteria of 0.03 carbon for “L” grade but also meets the mechanical properties of straight grade.  Straight grades have a higher tensile allowing for the use of a thinner plate than “L” grade plate.

Figure 5 presents an example of a separator made of stainless vessel. This 316 stainless separator is the first to be used offshore in place of a clad vessel. Since the temperature was low, the higher tensile allowed this item to be thinner, saving weight, and not require PWHT (Post Weld Heat Treatment), impact testing or special paint.

Summary:

In the last half century, the adoption of new technology in the manufacturing of fine grain steel plates, pipes and forgings has vastly improved the quality of the steels used in Oil and Gas Processing Equipment. Along with improvements in the welding processes used to construct Oil and Gas Processing Equipment, vessels, exchangers, piping and storage tanks are safer than ever before.

To learn more about similar cases and how to minimize operational problems, we suggest attending our ME43 (Mechanical Specification of Pressure Vessels and Heat Exchanges), PF49 (Troubleshooting Oil and Gas Facilities), PF42 (Separation Equipment Selection and Sizing), G4 (Gas Conditioning and Processing), and PF4 (Oil Production and Processing Facilities), courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

John R. Curry

Instructor and Consultant

References:

1. ASME Boiler and Pressure Vessel Code Section II, Part A., American Society of Mechanical Engineering, 1968.

There are different process configurations for adsorption dehydration systems. The most common arrangements are two-tower and three-tower configurations. One can find the details of the adsorption dehydration process and the descriptions of equipment in Chapter 18 of John M. Campbell textbook [1]. Figures 1 and 2 present a simplified process flow diagram for two-tower and a three-tower configurations, respectively. These units can reduce the water content of a gas stream to less than 0.1 ppmv.  The gas industry normally uses adsorption dehydration units upstream of a liquefied natural gas (LNG) plant or a deep natural gas liquid (NGL) extraction plant where the gas temperature reduces to less than -160 °C (-256 °F) and -100 °C (-148 °F), respectively. Removal of water content to this very low level is essential to prevent freezing.

In the two-tower system, while tower A is in the adsorption mode, tower B is regenerating. After tower A completes its adsorption cycle, it will switch to the regeneration mode and tower B starts its adsorption cycle. At any time one of the towers is adsorbing water while the other one is regenerating.

In the three-tower configuration, at any time two towers (e.g. A and B) in staggered parallel adsorption while the third tower (e.g. C) is regenerating. In this configuration, half of the feed gas flow rate is going through tower A and the other half passes through tower B as shown in the cycle chart embedded in Figure 2.

The PetroSkills’ May 2015 [2] and October 2015 tip of the month (TOTM) [3] discussed the efficient operation of molecular sieve dehydration units. Specifically, they discussed the benefits of standby time in the adsorption dehydration processes and impact of feed gas conditions.

This month’s TOTM compares the required size of major equipment for the two-tower system with the three-tower system. The comparison considers the following parameters.

1. Mass of Desiccant
2. Bed Diameter
3. Bed Height
4. Regeneration Gas Rate
5. Regeneration Heating Load
6. Regeneration Cooling Load
7. Regeneration Gas Heater Load
8. Tower wall thickness and mass

Table 1 presents descriptions of the two configurations.

Table 1. Tower configurations

This discussion assumes a gas volume rate of 2.83×10⁶ Sm³/d (100 MMSCFD) with externally insulated towers. It examines two feed conditions of:

a. 30°C (86°F) and 6.207 MPaa (900 psia)

• Estimated water content = 699 kg/10⁶ Sm³ (43.7 lb_m/MMSCF)
  • 2-Tower water load per tower per hour = 4 kg/h (182 lb_m/hr)
    • Water load per tower per cycle = 989 kg (2185 lb_m)
  • 3-Tower water load per tower per hour = 2 kg/h (91 lb_m/hr)
    • Water load per tower per cycle = 662 kg (1456 lb_m)

b. 40°C (104°F) and 8.0 MPaa (1160 psia)

• Estimated water content = 974 kg/10⁶ Sm³ (61 lb_m/MMSCF)
  • 2-Tower water load per tower per hour = 115 kg/h (254 lb_m/hr)
    • Water load per tower per cycle = 1378 kg (3050 lb_m)
  • 3-Tower water load per tower per hour = 5 kg/h (127 lb_m/hr)
    • Water load per tower per cycle = 920 kg (2033 lb_m)

The other specified parameters are:

1. Desiccant = Molecular Sieve Type 4A, 3.2 mm (1/8 in) diameter
2. Desiccant loading capacity, ΔX_new = 19 wt% (mass of water/100 mass of desiccant)
3. Desiccant life factor F_L = 0.6 (Based on 3 years and average performance)
4. Desiccant density = 705 kg/m³ (44 lb_m/ft³)
5. Desiccant heat capacity = 1.0 kJ/kg-°C (0.24 Btu/lb_m-°F)
6. Steel heat capacity = 0.5 kJ/kg-°C (0.12 Btu/lb_m-°F)
7. Feed gas relative density = 0.7
8. Regeneration dry lean gas relative density = 0.59
9. Regeneration gas pressure = 2.07 MPa (300 psia)
10. Regeneration gas temperature to the heater = Feed gas temperature
11. Regeneration gas temperature from heater = 288 °C (550 °F)
12. Final bed regeneration temperature prior to cooling = 260 °C (500 °F).

Calculation Results

Based on the procedure and steps of Chapter 18 [1], this TOTM utilized a revised version of PetroSkills/Campbell GCAP software [4] to perform all of the calculations.

Figure 3A shows variation of the required mass of desiccant per tower with the feed gas water content for two and three-tower systems.  As the feed water load increases, the required mass of desiccant increases for the specified and constant adsorption time. Note that the feed water load increases with increase in temperature and decrease in pressure. The water load per tower is a function of the feed water content, adsorption time and gas flow rate through the tower.

The adsorption cycle times chosen for the two-tower and three-tower systems (12 hours and 16 hours, respectively) result in essentially the same total mass of water being adsorbed by both configurations. The total required mass of desiccant for the two-tower and three-tower systems are different because the length of mass transfer zone for the 3-tower system is lower than for the 2-tower system. Since desiccants are sold in 300 lb_m (136.1 kg) increment, the mass of desiccant in each tower was rounded up to the next 300 lb_m (136.1 kg).

The calculated total mass of desiccants for the two-tower system are 19323 and 25855 kg (42600 and 57000 lb_m) for the feed with the lower and higher water content, respectively. The corresponding desiccant masses for the three-tower system are 18779 and 25719 kg (41400 and 56700 lb_m).  Similarly, Figure 3B presents the required mass of steel per tower as a function of feed gas water content and number of towers.

The mass of the desiccant in the tower and pressure drop criteria establish the tower diameter and height of desiccant. This is a trial and error calculation. Figure 4 shows variation of the minimum required bed diameter with the feed gas water content and number of towers. Since the gas flow rate per tower of the two-tower system is two times higher than the gas flow rate per tower in the three-tower system, the diameter of each tower in the two-tower system must be larger to meet pressure drop criteria of less than 41 kPa (6 psi). Figure 4 indicates that the limiting factor is the pressure drop and the feed gas water content has small effect on the bed diameter. The calculated superficial gas velocity for all cases ranged from 0.10 to 0.15 m/s (20 to 30 ft/min).

After determining the bed diameter, one can calculate the desiccant height from the mass of desiccant, bed diameter, and the desiccant density.  Figure 5 shows the variation of the minimum desiccant height with the feed gas water content and number of towers.  This figure indicates that feed with higher water content requires a taller bed and feed with lower water load requires shorter bed height. Since the diameter in two tower system was larger its height is shorter.

Figure 6 shows the variation of the regeneration gas requirement with the feed gas water content while maintaining constant heating and cooling times.  This figure indicates that higher feed gas water content require more regeneration gas. However, the required regeneration gas rate is practically the same for the two-tower and three-tower systems under evaluation.

Similarly, Figures 7, 8, and 9 show the variation of the required heating load, cooling load, and regeneration gas heater load, with the feed gas water content and number of towers. Since the towers in the two-tower system are larger than the towers in the three-tower system, the heating and cooling loads are larger for the towers in the two-tower systems (figures 7 and 8). However, Figure 9 indicates that the required regeneration heater loads are almost the same for the two-tower and three-tower systems.

Summary:

Water content of the feed gas affected by temperature, pressure and flow rate is the key factor in sizing and operation of adsorption dehydration system. Higher water load requires a larger size bed, higher heating and cooling loads and higher rate of regeneration gas.

In the cases evaluated in this Tip of the Month, the tower diameters in the two-tower systems are larger but their heights are shorter. The mass of desiccants and mass of steel per tower in the two-tower system are larger than in the three-tower system. Therefore the heating and cooling loads are larger for the towers in the two-tower system. The regeneration gas heater loads are almost the same for two and three-tower system.

To learn more about similar cases and how to minimize operational problems, we suggest attending our PF49 (Troubleshooting Oil and Gas Facilities), PF42 (Separation Equipment Selection and Sizing), G4 (Gas Conditioning and Processing), G5 (Gas Conditioning and Processing – Special), and PF4 (Oil Production and Processing Facilities), courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

Dr. Mahmood Moshfeghian

References:

1. Campbell, J.M., Gas Conditioning and Processing, Volume 2: The Equipment Modules, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.
2. Malino, H. M., http://www.jmcampbell.com/tip-of-the-month/2015/05/benefits-of-standby-time-in-adsorption-dehydration-process/
3. Moshfeghian, M., http://www.jmcampbell.com/tip-of-the-month/2015/10/what-is-the-impact-of-feed-gas-conditions-on-the-adsorption-dehydration-system/
4. GCAP Version 9.1.1, Gas Conditioning and Processing Software, Editor Moshfeghian, M., PetroSkills/Campbell, Norman, Oklahoma, 2015.

Download PDF Version

Adsorption dehydration units can reduce the water content of a gas stream to less than 0.1 ppmv.  The gas industry normally uses adsorption dehydration units upstream of a liquefied natural gas (LNG) plant or a deep natural gas liquid (NGL) extraction plant where the gas temperature reduces to less than -160 °C (-256 °F) and -100 °C (-148 °F), respectively. Removal of water content to this very low level is essential to prevent freezing. One can find the details of the adsorption dehydration process and the descriptions of equipment in Chapter 18 of John M. Campbell textbook [1].

The PetroSkills’ May 2015 tip of the month (TOTM) [2] discussed the efficient operation of molecular sieve dehydration units. Specifically, it discussed the benefits of standby time in the adsorption dehydration processes.

This month’s TOTM discusses feed gas flow rate, pressure, and temperature effect(s) on the followings:

1. Mass of Desiccant
2. Bed Diameter
3. Bed Height
4. Regeneration Gas Rate
5. Regeneration Heating Load
6. Regeneration Cooling Load
7. Regeneration Gas Heater Load

This discussion spans gas volume rates from 2.83×10⁶ to 11.32×10⁶ Sm³/d (100 to 400 MMSCFD. It examines this range at two temperatures:  30°C (86°F) and 40°C (104°F) with each temperature evaluated at two adsorption pressures: 6.207 MPaa (900 psia) and 8.0 MPaa (1160 psia).

Figure 1 shows a simplified process flow diagram for a 3-tower system [1]. In this TOTM, the specified parameters are:

1. Desiccant = Molecular Sieve Type 4A, 3.2 mm (1/8 in) diameter
2. No of towers = 3
3. No of towers on-stream = 2
4. Adsorption time = 16 hours
5. Step time (same as total regeneration time) = 8 hours
6. Stand-by time = 0 hours
7. Equalization time = 0.5 hours
8. Heating time = 4.88 hours
9. Cooling time = 2.62 hours
10. Desiccant loading capacity, ΔX_new = 19 wt% (mass of water/100 mass of desiccant)
11. Desiccant life factor F_L = 0.6 (Based on 3 years and average performance)
12. Desiccant relative density = 0.705
13. Desiccant heat capacity = 1.0 kJ/kg-°C (0.24 Btu/lb_m-°F)
14. Steel heat capacity = 0.5 kJ/kg-°C (0.12 Btu/lb_m-°F)

Calculation Results

Based on the procedure and steps of Chapter 18 [1], this TOTM utilized a revised version of PetroSkills/Campbell GCAP software [3] to perform all of the calculations.

Figure 2 shows variation of the required mass of desiccant with the feed gas rate, pressure and temperature.  As the water load increases, the required mass of desiccant increases for a constant adsorption time. Note that the feed water load increases with increase in feed gas rate and temperature and decrease in pressure. The feed with higher temperature and lower pressure requires more mass of desiccant. For the same reason, the feed with lower temperature and higher pressure requires less mass of desiccant.

The mass of the desiccant in the tower and pressure drop criteria establish the tower diameter and height of desiccant. This is a trial and error calculation. Figure 3 shows variation of the minimum bed diameter with the feed gas rate, pressure and temperature. This figure indicates that feed with higher water load (higher temperature and lower pressure) needs a bed with larger diameter and feed with lower water load (lower temperature and higher pressure) requires a smaller diameter.

One can calculate the desiccant height from the mass of desiccant, bed diameter, and desiccant density.  Figure 4 shows the variation of the minimum desiccant height with the feed gas rate, pressure and temperature.  Similarly, this figure indicates that feed with higher water load (higher temperature and lower pressure) requires a taller beds and feed with lower water load (lower temperature and higher pressure) requires shorter bed heights.

For all cases, the lean and dry regeneration gas had a relative density of 0.59 and pressure of 2.069 MPaa (300 psia) but its temperature was the same as the feed gas temperature. Its temperature from heater was 532.2 °C (550 °F) and the final bed regeneration temperature prior to cooling was 288 °C (500 °F).

Figure 5 shows the variation of the regeneration gas requirement with the feed gas rate, pressure and temperature while maintaining constant heating and cooling times.  This figure indicates that higher water loads require more regeneration gas.  Similarly, Figures 6, 7, and 8 show the variation of the required heating load, cooling load, and regeneration gas heater load, respectively, with the feed gas rate, pressure and temperature.

Summary:

Water content of the feed gas affected by temperature, pressure and flow rate is the key factor in sizing and operation of adsorption dehydration system. Higher water load requires a larger size bed, higher heating and cooling loads and higher rate of regeneration gas.

To learn more about similar cases and how to minimize operational problems, we suggest attending our PF49 (Troubleshooting Oil and Gas Facilities), PF42 (Separation Equipment Selection and Sizing), G4 (Gas Conditioning and Processing), G5 (Gas Conditioning and Processing – Special), and PF4 (Oil Production and Processing Facilities), courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

Dr. Mahmood Moshfeghian

References:

1. Campbell, J.M., Gas Conditioning and Processing, Volume 2: The Equipment Modules, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.
2. Malino, H. M., http://www.jmcampbell.com/tip-of-the-month/2015/05/benefits-of-standby-time-in-adsorption-dehydration-process/
3. GCAP Version 9.1.1, Gas Conditioning and Processing Software, Editor Moshfeghian, M., PetroSkills/Campbell, Norman, Oklahoma, 2015.

Download PDF Version

Download PDF Version

In the December 2014 tip of the month (TOTM) [1], we discussed troubleshooting of gas-liquid separators for removal of liquids from the gas stream leaving the separator. There are two methods for sizing gas-liquid separators: 1. Droplet settling theory method, 2. Souders-Brown approach. Historically the Souders-Brown equation has been employed as it can provide reasonable results and is easy to use, but has shortcomings in terms of quantifying separator performance.  References [2-4] provide the details on the droplet settling theory methods which can be used to more accurately quantify separator performance.  The Souders-Brown method is limited in that it is based on the average droplet size, but cannot quantify the amount of liquid droplets exiting the gas gravity section.

In this TOTM, we will focus on the application of Souders-Brown approach in gas-liquid separators and present diagram, simple correlations and tables to estimate the Souders-Brown equation constant, K_S (the so called sizing parameter). We will consider both vertical and horizontal gas-liquid separators. Knowing the actual gas flow rate through the vessel, one can use K_S parameter to determine the maximum allowable gas velocity through the vessel and determine the required separator diameter. One can also use the appropriate value of K_S to size the mist extractor in the vessel. The performance of a gas-liquid separator is highly dependent on the value of K_S; therefore, the choice of appropriate  K_S –values is important.

Gas Gravity Separation Section

The gas gravity separation section of a separator has two main functions:

1. Reduction of entrained liquid load not removed by the inlet device
2. Improvement / straightening of the gas velocity profile.

Most mist extractors have limitations on the amount of entrained liquid droplets that can be efficiently removed from the gas, thus the importance of the gas gravity section to remove the liquids to an acceptable level upstream of the mist extractor.  This is particularly important for separators handling higher liquid loads. For scrubber applications with low liquid loadings,  the K_S –values will be primarily dependent on the mist extractor type, and the gas gravity separation section becomes less important. For the higher liquid load applications, i.e. conventional separators, there are two approaches for sizing the gravity separation section to remove liquid droplets from the gas:

1. The Souders-Brown approach (K_s Method)
2. Droplet settling theory

The Souders-Brown Approach

If we consider a spherical liquid droplet with a diameter of D_P in the gas phase two forces as shown in Figure 1 act on it. The drag force, F_D, is exerted by flow of gas and gravity force, F_G, is exerted by the weight of droplet. The drag force acts to entrain the liquid droplet while the gravity force acts to pull it down and separating it from the gas phase.

Assuming plug flow with no eddies or disturbances, a single droplet and ignoring the end effect, at equilibrium (free fall or terminal velocity), these two forces are equal.

       (1)

As presented in the Appendix, substitution of expressions for the drag and gravity forces in Equation 1, the maximum allowable gas velocity, V_Gmax, which prevents entrainment of liquid is obtained.

Equation 2 is called Souders-Brown [6] equation and K_S is referred to as the design or sizing parameter. The terms ρ_G and ρ_L are the gas phase and liquid phase densities, respectively.

Once the maximum gas velocity, V_Gmax, through the vessel is determined by Equation 2, one can calculate the minimum vessel diameter, D_min by Equation 3.

Where:

F_G        = Fraction of cross section area available for gas flow (F_G = 1 for vertical separators and is a function of liquid height for horizontal separators)

q_a         = Gas flow rate at the actual flowing condition

The Design Parameter, K_S

The design parameter, K_S, in the Souders-Brown equation is an empirical parameter and is a key factor for the sizing the gas-liquid separators vessel diameter as well as for determination of the mist extractor diameter. Its value depends on several factors including:

• Pressure
• Fluid properties (note temperature has a large impact on fluid properties)
• Separator geometry
  • Vessel length and the liquid level (for horizontal separators)
• Steadiness of flow
• Inlet device design and performance
• Relative amounts of gas and liquid
• Most importantly – mist extractor type and design (e.g. mesh pad, vane pack, multi–cyclone)

There are several sources that one can look up the K_S –values for different applications.  In the following sections, we will discuss three sources.

A. API 12 J

The API 12J [7] recommends ranges of K_S –values for vertical and horizontal gas-liquid separators. These values are presented in Table 1. The equivalent of API 12J for the North Sea region is NORSOK P-100.

Table 1. API 12 J recommended range of K_S –values for vertical and horizontal separators [7]

Per API 12J, “the maximum allowable superficial velocity, calculated form the above factors, is for separators normally having a wire mesh mist extractor. This rate should allow all liquid droplets larger than 10 microns to settle out of the gas. The maximum allowable superficial velocity or other design criteria should be considered for other type mist extractor. Mist extractor manufacturer’s recommended minimum distances upstream and downstream of the wire mesh between gas inlet and outlet nozzles should be provided for full utilization of the mist extractor. These values assume separators are equipped with standard mesh pad mist extractors” [7].

B. Campbell Book

The K_s method, Equation 2, is an empirical approach to estimate the maximum allowable gas velocity to achieve a desired droplet separation. For vertical separators with no mist extractor devices, Chap 11, Vol 2 of the Gas Conditioning and Processing book presents K_S as a function of pressure and liquid droplet size [5]. This dependency of K_S on pressure and droplet size is presented in Figure 2 [5]. Note for each droplet size a range of K_S –values are given for a specified pressure.

For horizontal separators, the sizing depends on (in addition to the droplet size, density of gas and liquid phases, and gas velocity) separator effective length, L_e, and the depth available for gas flow, h_G, (i.e. liquid level) in the separators.

Sizing of the horizontal separators are more complicated. Referring to Figure 3, the effective L_e may be defined in terms of separator actual length and diameter like L_e=L-D. Therefore, the Souders-Brown parameter for horizontal separators, K_SH, can be estimated in by Equation 4 in terms of K_SV (read from Figure 2) for vertical separator [3].

If the calculated value of K_SH by Equation 4 is greater than the maximum value of 0.7 ft/sec (0.21 m/s), it should be set equal to this maximum value.

The horizontal separator sizing is a trial-and-error procedure. Normally, the L_e/D and h_g/D (or h_L/D) are assumed and K_SH, V_gmax, D are calculated by Equations 4, 2, and 3, respectively. The effective length and actual lengths are calculated by Equation 5.

Where:

D         = Diameter

F_L        = Fraction of cross section area occupied by liquid (function of liquid height in horizontal separator)

q_L         = Liquid actual volume flow rate

t           = Residence time per API 12J [7]

If the calculated L/D is outside of its recommended range (normally 3 < L/D < 6), the liquid height in the vessel is changed and the calculations are repeated. For detail of calculations procedure refer to chapter 11 of reference [5].

C. K_S Correlations

The curves for different droplet sizes shown in Figure 2 are fitted to a 3^rd order polynomial (for droplet sizes of 100, 150, and 300 microns). The correlation is in the form of Equation 6 and its regressed coefficients a, b, c, and d are presented in Tables 2A and 2B for field (FPS) and System International (SI) units, respectively.

In Table 2, each droplet size in micron (µ) is preceded by letter L or U representing the lower and upper curve, respectively. The pressure is in psi and K_S is in ft/sec for FPS (kPa and m/s in SI).

The last row of Table 2 provides the average absolute percent deviation (AAPD) of the predicted K_S by the proposed correlation from the corresponding values of Figure 2 values.

Table 2A (FPS). Regressed coefficients for Equation 6 (P in psi and K_S in ft/sec)
Droplet Size: 100 – 300 microns

Table 2B (SI). Regressed coefficients for Equation 6 (P in kPa and K_S in m/s in SI).
Droplet Size: 100 – 300 microns

The two curves for 500 micron droplet size in Figure 2 were divided into 4 and 2 segments based on pressure range for the lower and upper curves, respectively. Each segment was fitted to a linear equation in the form of Equation 7 and its regressed coefficients e and f are presented in Tables 3A and 3B for FPS and SI units, respectively.

Table 3A (FPS). Regressed coefficients for Equation 7 (P in psi and K_S in ft/sec)
Droplet Size: 500 microns

Table 3B (SI). Regressed coefficients for Equation 7 (P in kPa and K_S in m/s in SI).
Droplet Size: 500 microns

D. Mist Extractors

The mist extractor is the final gas cleaning device in a conventional separator.  The selection, and design to a large degree, determine the amount of liquid carryover remaining in the gas phase.  The most common types include wire mesh pads (“mesh pads”), vane-type (vane “packs”) and axial flow demisting cyclones.  Figure 4 shows the location and function of a typical mist extractor in a vertical separator.

Mist extractor capacity is defined by the gas velocity at which re-entrainment of the liquid collected in the device becomes appreciable.  This is typically characterized by a K_S –value, as shown in Equation 2.  Mesh pads are the most common type of mist extractors used in vertical separator applications.  The primary separation mechanism is liquid impingement onto the wires, followed by coalescence into droplets large enough to disengage from the mesh pad.  References [1-5] provide mesh pad examples.  Table 4 provides a summary of mesh pad characteristics and performance parameters.

Table 4. Mesh pads K_S and performance parameters [3, 5, 8]

Notes:

• Flow direction is vertical (upflow).
• Assume mesh pad K_S – value decline with pressure as shown in Table 5. Table 5 was originally developed for mesh pads, but is used as an approximation for other mist extractor types. [9].
• If liquid loads reaching the mesh pad exceed the values given in Table 4, assume capacity (K_S) decreases by 10% per 42 L/min/m² (1 gal/min/ft²). [2-4].
• These parameters are approximate.

Table 5. Mesh pad K_S deration factors as a function of pressure [5]

Vane packs, like mesh pads, capture droplets primarily by inertial impaction.  The vane bend angles force the gas to change direction while the higher density liquid droplets tend to travel in a straight-line path, and impact the surface of the vane where they are collected and removed from the gas flow. Table 6 provides vane pack performance characteristics [3, 5, 8].

In the case of demisting cyclones, the vendor should be consulted in regards to performance for the current operations of interest.

Table 6. Typical vane-pack characteristics [3, 5, 8]

Notes:

1. Assume vane-pack K_S – value decline with pressure as shown in Table 5.
2. If liquid loads reaching the vane pack exceed the values given in Table 2, assume capacity K_S decreases by 10% per 42 L/min/m² (1 gal/min/ft²). [2-4].
3. These parameters are approximate only. The vane-pack manufacturer should be contacted for specific information.

Conclusions: 

We focused on the application of Souders-Brown Approach (SBA) in gas-liquid separators and presented diagram, simple correlations and tables to estimate the SBA design parameter, K_S.

• The SBA can provide reasonable results and is easy to use.
• The SBA is limited in that it is based on the average droplet size, but cannot quantify the amount of liquid droplets exiting the gas gravity section and mist extractor section.
• In a future TOTM we will discuss the droplet settling theory methods which can be used to more accurately quantify separator performance.
• Sizing of three-phase gas-liquid hydrocarbon-liquid water separators are more complicated and will be discussed in another TOTM.

To learn more about similar cases and how to minimize operational problems, we suggest attending our PF49 (Troubleshooting Oil and Gas Facilities), PF42 (Separation Equipment Selection and Sizing), G4 (Gas Conditioning and Processing), G5 (Gas Conditioning and Processing – Special), and PF4 (Oil Production and Processing Facilities), courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

Dr. Mahmood Moshfeghian

References:

1. Snow–McGregor, K., http://www.jmcampbell.com/tip-of-the-month/2014/12/troubleshooting-gas-liquid-separators-removal-of-liquids-from-the-gas/
2. Bothamley, M., “Gas-Liquid Separators – Quantifying Separation Performance Part 1,” SPE Oil and Gas Facilities, pp. 22 – 29, Aug. 2013.
3. Bothamley, M., “Gas-Liquid Separators – Quantifying Separation Performance Part 2,” SPE Oil and Gas Facilities, pp. 35 – 47, Oct. 2013.
4. Bothamley, M., “Gas-Liquid Separators – Quantifying Separation Performance Part 3,” SPE Oil and Gas Facilities, pp. 34 – 47, Dec. 2013.
5. Campbell, J.M., Gas Conditioning and Processing, Volume 2: The Equipment Modules, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.
6. Souders, M. and Brown, G. G., “Design of Fractionating Columns-Entrainment and Capacity,” Industrial and Engineering Chemistry, Volume 26, Issue 1, p 98-103, 1934.
7. American Petroleum Institute, 12J, Specification for Oil and Gas Separators, 8th Edition, October, 2008.
8. PF-49, Troubleshooting Oil and Gas Processing Facilities, Bothamley, M., 2014, © PetroSkills, LLC. All Rights reserved.
9. Fabian, P., Cusack, R., Hennessey, P., Neuman, M., “Demystifying the Selection of Mist Eliminators, Part 1: The Basics,” Chem Eng 11 (11), pp. 148 – 156, 1993.

 

Appendix

Derivation of the Souders-Brown and Stokes’ Law Equations

If we consider a spherical liquid droplet with a diameter of, D_P, in the gas phase two forces as shown in Figure 1 act on it. The drag force, F_D, is exerted by flow of gas and gravity force, F_G, is exerted by weight of droplet. The drag force acts to entrain the liquid droplet while the gravity force acts to pull it down.

At equilibrium, these two forces are equal.

The drag force is expressed as:

The droplet projected area, A_P, is defined by:

The gravity force, F_G, is defined

The volume of spherical droplet, V_D, is calculated by

Substitution of Equations 3 and 4 into Equation 1 and solving for the gas maximum velocity,

For practical applications, the first term on the right hand side is replaced by K_S

Therefore, the maximum gas velocity which prevents entrainment of liquid is obtained.

Equation 6 is called Souder-Brown equation and K_S is referred to as the design parameter.

Where:

A_P        = Project area of droplet

C_D        = Drag coefficient

g          = Acceleration of gravity

g_C        = Conversion factor

V          = Gas velocity

V_P        = Volume of droplet

ρ_G        = Gas density

ρ_L         = Liquid density

Once the maximum, V_Gmax, gas velocity through the vessel is determined by Equation 6, one can calculate the required minimum cross sectional area of vessel for gas flow by the following equation.

Solving for the minimum vessel diameter, D_min.

Where:

F_G        = Fraction of cross sectional area available for gas flow (F_G = 1 for vertical separators and it is a function of liquid height for horizontal separators)

q_a         = Gas flow rate at the actual flowing condition

The drag coefficient, C_D, is a function of Reynolds number, Re=(D_PVρ_G)/µ_G. For Stokes’ law Re <≈2

Substitution of C_D from Equation 9 into Equation 4 gives liquid droplet terminal velocity, V_T, in the gas phase based on the Stokes’ law.

Similarly, the terminal velocity for the other flow regimes like Intermediate and Newton can be derived based on their corresponding expressions for the drag coefficients [3].

Download PDF Version

In the past Tips of the Month (TOTM), we discussed the phase behavior and water content of lean sweet, sour natural gases and acid gases–water systems. Specifically, in the November 2007 [1], February 2014 [2], and September 2014 [3] Tips of the Month (TOTM), we discussed the phase behavior of water-saturated sour gases and acid gases. We also demonstrated the accuracy of shortcut and rigorous calculation methods. In the April 2015 TOTM [4] we introduced correlation and graphs to estimate the water content of sour gases.

In this TOTM, we will study the effect of relative density (Specific Gravity, SG) on the saturated water content of sweet natural gases. The results of this study include the water content of sweet natural gases as a function of relative density in the range of 0.60 to 0.80.  Four temperatures of 4.4, 23.9, 37.8 and 149 °C (40, 75, 100, and 300 °F) were considered. For each temperature, the saturated water content was calculated for pressures of 1724, 3448, 6897, and 13 793 kPaa (250, 500, 1000 and 2000 psia).

Water Content Calculation

The dry gas compositions of the four mixtures studied in this study are presented in Table 1. The Soave-Redlich-Kwong equation of state (SRK EOS) [5] in the ProMax [6] was used to predict the water content of these gas mixtures at different pressures and temperatures. A simplified process diagram for this study is presented in Figure 1. The feed dry-gas at the specified pressure and temperature was saturated with water first and passed through a separator. The water content and specific gravity of vapor leaving the separator were recoded.

Results and Discussion

Figure 2 through 5 present the saturated water content of sweet natural gases as a function of the dry gas  relative density and pressures of 1724 to 13793 kPaa (250 to 2000 psia) for temperatures of 4.4, 23.9, 37.8 and 149 °C (40, 75, 100, and 300 °F), respectively.

Table 1. The composition of relative density (Specific Gravity, SG) of the four gas mixture

While the results presented in the above diagrams are in agreement with Figure 6.1 of Reference [7], it is not in agreement with the suggested correction factor in the inset of Figure 20-3 of the GPSA data book [8] and the results of Reference [9]. For the gases with relative density of 0.6 up to 0.8, the GPSA figure indicates a correction factor of 1 to about 0.97 should be multiplied by the water content of saturated water content of sweet gas with relative density of 0.6. This study indicates that for the range of 0.6 to 0.8 relative density, the water content is not a function of the hydrocarbon gas composition. To verify the result of this study, the saturated water content of several gas mixtures were compared with the results of EzThermo [10] software. The SRK EOS in the EzThermo software was developed and regressed to predict the properties of sweet synthetic natural gas and natural gas compositions [11]. The same is also valid for the SRK EOS in ProMax. The comparison of these two software is presented in Table 2. This table indicates an excellent agreement between these two software.

Table 2. Comparison of the ProMax [6] and EzThermo [10] predicted saturated water content results at 37.8 °C (100 °F)

Conclusions

Figures 2 through 5 cover wide ranges of pressures and temperatures commonly encountered in the gas processing operation. The analysis of Figures 2 through 5 indicates that the gas relative density has minimal effect on the saturated water content of sweet natural gases. This conclusion is valid for the following ranges:

1. Sweet gas relative density in the range of 0.6 to 0.8
2. Temperature range of 4.4 to 149 °C (40 to 300 °F)
3. Pressure range of 1724 to 13793 kPaa (250 to 2000 psia)

Additional experimental water content data are being taken and analyzed by the GPA Research Committee to update the relative density correction factor as presented in [8].  From the data that are currently available, it appears that this correction is minimal if any for gas compositions, and temperature and pressure ranges that typically occur in the oil and gas facilities. This will be confirmed once the results from the GPA Research Committee have been published.

To learn more about similar cases and how to minimize operational problems, we suggest attending our G4 (Gas Conditioning and Processing), G5 (Gas Conditioning and Processing – Special), G6 (Gas Treating and Sulfur Recovery), PF49 (Troubleshooting Oil and Gas Facilities),  and PF81 (CO₂ Surface Facilities) courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

Dr. Mahmood Moshfeghian

Reference:

1. http://www.jmcampbell.com/tip-of-the-month/2014/09/lean-sweet-natural-gas-water-content-correlation/
2. http://www.jmcampbell.com/tip-of-the-month/2007/11/water-sour-natural-gas-phase-behavior/
3. http://www.jmcampbell.com/tip-of-the-month/2014/09/lean-sweet-natural-gas-water-content-correlation/
4. http://www.jmcampbell.com/tip-of-the-month/2014/02/acid-gas-water-content/
5. Soave, G., Chem. Eng. Sci., Vol. 27, pp. 1197-1203, 1972.
6. ProMax 3.2, Bryan Research and Engineering, Inc., Bryan, Texas, 2014.
7. Campbell, J.M., “Gas conditioning and Processing, Vol. 1: The Basic Principles”, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.
8. GPSA Engineering Data Book, Section 20, Volume 2, 13^th Edition, Gas Processors and Suppliers Association, Tulsa, Oklahoma, 2012.
9. Maddox, R.N., L.L. Lilly, M. Moshfeghian, and E. Elizondo, “Estimating Water Content of Sour Natural Gas Mixtures”, Laurence Reid Gas Conditioning Conference, Norman, OK, Mar., 1988.
10. EzThermo, Moshfeghian, M. and R.N. Maddox, 2015.
11. GPA Research Report RR-42, Predicting Synthetic Gas and Natural Gas Thermodynamic Properties Using a Modified Soave Redlich Kwong Equation of State, Oklahoma State University, August 1980

In the November 2011 tip of the month (TOTM) we presented the compressor calculations of a case study. We compared the rigorous method results with the values from the shortcut methods. The rigorous method was based on an equation of state like the Soave-Redlich-Kwong (SRK) for calculating the required enthalpies and entropies. The enthalpies and entropies are used to determine the power requirement and the discharge temperatures.  The results indicated that the accuracy of the shortcut method is sensitive to the value of ideal gas state heat capacity ratio, k.

From a calculation viewpoint alone, the power calculation is particularly sensitive to the specification of mass flow rate, suction temperature and pressure, and discharge temperature and pressure. A compressor is going to operate under varying values of the variables affecting its performance. Thus the most difficult part of a compressor calculation is specification of a reasonable range for each variable and not the calculation itself. Reference [1] emphasizes that using a single value for each variable is not the correct way to evaluate a compression system.

Normally, the thermodynamic calculations are performed for an ideal (reversible) process. The results of a reversible process are then adapted to the real world through the use of a thermodynamic efficiency. In the compression process there are three ideal processes that can be visualized: 1) an isothermal process (PV¹=C₁), 2) an isentropic process (PV^k=C₂) and 3) a polytropic process (PVⁿ=C₃). Any one of these processes can be used suitably as a basis for evaluating compression power requirements by either hand or computer calculation. The isothermal process, however, is seldom used as a basis because the normal industrial compression process is not even approximately carried out at constant temperature.

Note that Dresser Rand is doing quite a lot of work with “Near constant temperature” compression especially for CO₂ compression from vent stacks. For detail refere to:

http://www.nist.gov/pml/high_megawatt/upload/6_1-Approved-Moore.pdf

In this TOTM, we will demonstrate how to determine the efficiency of a compressor from measured flow rate, composition, suction and discharge temperatures and pressures. A rigorous calculation based on an equation of state and a shortcut method are considered and the results are compared.

Compress Efficiency

Compressor efficiencies vary with compressor type, size, and throughput. They can only be determined (afterward) by a compressor test, although compressor manufacturers can usually provide good estimates. For planning purposes, reference [2] suggests the following values for the overall efficiencies:

Table 1. Overall Compressor Efficiencies [2]

Compressor Type	Efficiency, η
Centrifugal	0.70 – 0.85
High Speed Reciprocating	0.72 – 0.85
Low Speed Reciprocating	0.75 – 0.90
Rotary Screw	0.65 – 0.75

Reference [2] indicates that these overall efficiencies include gas friction within the compressor, the mechanical losses (bearings, seals, gear-box, etc.), and gear-box losses. The mechanical efficiency varies with compressor size and type, but 95% is a useful planning number. When calculating the compressor head and discharge temperature the efficiency used will be isentropic or polytropic (isentropic efficiency is sometimes called adiabatic efficiency). Adding 3-4 % efficiency (mechanical losses) to the overall efficiencies in Table 1 will generally give a good estimate of the thermodynamic efficiency [2].

To evaluate the performance of an existing compressor, the objective is to calculate the compressor efficiency (η) and power requirement.

Known and measured properties are:

a. Standard condition gas volume flow rate (q_S) or gas mass rate ()

b. Gas composition (z_i)

c. Suction pressure (P₁) and  temperature (T₁)

d. Discharge pressure (P₂) and temperature (T₂)

Estimating Efficiency – Rigorous Method

The heart of any commercial process flow simulation software is an equation of state. Due to their simplicity and relative accuracy, a cubic EOS such as Soave Redlich-Kwong (SRK) [3] or Peng-Robinson [4] is used. These equations are used to calculate Vapor-Liquid-Equilibria (VLE), enthalpy (h), and entropy (s). With proper binary interaction coefficients, the process simulation results of these two equations are practically the same. Therefore, only the SRK is used in this work.

The isentropic efficiency is defined by

Where:

_{ηIsen = Isentropic efficiency}

h₁ = Suction enthalpy calculated at P₁, T₁, and composition (z_i)

h₂ = Discharge enthalpy calculated at P₂, T₂, and composition (z_i)

h_2Isen = Isentropic discharge enthalpy at P₂ (or T₂), S_2Isen =S₁, and composition (z_i)

 = Mass flow rate

The computation compressor efficiency or power involves two steps

1. Determination of the ideal or isentropic (reversible and adiabatic) enthalpy change (h_2Isen-h₁) of the compression process.

2. Determination of the actual enthalpy change (h₂-h₁).

The step-by-step calculation based on an EOS:

a. Assume steady state, i.e.  

b. Assume the feed composition remain unchanged

c. Calculate suction enthalpy h₁=f(P₁, T₁, and z_i) and entropy s₁=f(P₁, T₁, and z_i) by EOS

d. Assume isentropic process and set s_2Isen = f (P₂, T_2Isen, z_i) = s₁ = f (P₁, T₁, z_i).

e. Calculate the ideal enthalpy (h_2Isen) at discharge condition for known z_i, T₂ (or P₂) and s_2Isen.

f. Calculate the actual enthalpy (h₂) at discharge condition for known z_i, T₂ and P₂.

g. Calculate isentropic efficiency by Equation 1: µ_Isen = (h_2Isen – h₁)/(h₂ – h₁)

h. Calculate power by Equation 2: 

Estimating Efficiency – Shortcut Method

The isentropic path exponent (k) or ideal gas heat capacity ratio (k=C_P/C_V) can be calculated by the correlation presented in the May 2013 TOTM:

Where:

T = Temperature, K (°R)

 = Gas relative density; ratio of gas molecular weight to air molecular weight

A = 0.000272 (0.000151)

The actual discharge temperature based on an isentropic path can be estimated by

Solving for the isentropic efficiency,

Similarly, the actual discharge temperature based on a polytropic path can be estimated (η_Poly) by:

The isentropic head is calculated by

Similarly, the polytropic head is calculated by

For an isentropic (reversible and adiabatic) process the power is calculated by

Or for a polytropic process the power is calculated by

Alternatively:

Where:

Head = Compressor head, m (ft)

Power = Compressor power, kW (HP)

R = Universal gas constant, 848 kg-m/(kmol-K) or (1545 ft-lb_f/(lbmol-°R))

P_S = Standard condition pressure, kPa (psia)

P₁ = Suction pressure, kPa (psia)

P₂ = Discharge pressure, kPa (psia)

T_S = Standard condition temperature, K (°R)

T₁ = Suction temperature, K (°R)

T₂ = Discharge temperature, K (°R)

q_S = Gas volumetric rate at the standard condition, Sm³/d (scf/day)

Za = Average gas compressibility factor = (Z₁+Z₂)/2

Z₁ = Gas compressibility factor at the suction condition

Z₂ = Gas compressibility factor at the discharge condition

MW = Gas molecular weight

The power calculation should be made per stage of compression and then summed for all stages connected to a single driver.

The step-by-step calculation for shortcut method

a. Calculate the isentropic exponent (k) by Equation 3 using the average temperature defined by T = (T₁+3T₂)/4. This form of average temperature was defined to obtain better match between the rigorous and shortcut method results.

b. Calculate the isentropic efficiency (η_Isen) by Equation 5.

c. Calculate the polytropic coefficient (n) by Equation 7.

d. Calculate the polytropic efficiency (η_Poly) by Equation 8.

e. Calculate the isentropic and polytropic heads by Equations 9 and 10, respectively.

f. Calculate the required power per stage by either Equation 11 or 12.

Case Study

A natural gas mixture is compressed using a three-stage centrifugal compressor. The process flow diagram is shown in Figure 1. For each stage, the measured pressure, and temperature are presented in Table 1. The measured feed composition, flowrates, and calculated molecular weight and relative density are presented in Table 2.

Table 1. Measured temperature and pressure for the three stages of compression

Table 2. Gas analysis and flow rate for the three stages of compression

Results and Discussions

The process flow diagram shown in Figure 1 was simulated by ProMax software [5] to perform the rigorous calculations using the SRK EOS. The program calculated polytropic and isentropic efficiencies, heads, and compression power. The program also calculated the isentropic path exponent (k), and polytropic path exponent (n). These calculated results are presented in Table 2 for all three stages under SRK headingings. The calculations performed by ProMax are very similar to the step-by-step of a through h described in the rigorous section. Table 2 also presents the shortcut caculation results for the corresponding values under the shortcut heading. The shortcut calculations are based on the step-by-step of a through f described in the shortcut method section. The error percent between the rigrous method and the shortcut methods for each stage are presented in Table 2, too. Table 2 indicates that excellent agreements are obtained for stages 1 and 2. However, larger deviations are obseved for the isetropic and polytropic exponents of stage 3 due to high pressure operation which deviated too far from ideal gas state conditions.

Table 3. Summary of the rigorous and shortcut calculated results

Conclusions

Table 2 indicates that there are good agreements between the shortcut and the rigorous results. The differences between the rigorous and shortcut method results for facilities calculations and planning purposes are negligible. For stage 3, due to high-pressure operation and deviating too far from the ideal gas state condition, a larger error is observed for the isentropic exponent (k).

The calculated isentropic exponent (k) in the ProMax [5] is not the ideal gas state heat capacity (C_P/C_V) ratio.  It is the value of the isentropic exponent that is required to yield an isentropic path from inlet to outlet.  Its value is calculated as an integration of that path.  Thus it is somewhat of an “average” value representing the true isentropic path.  For ideal gases, the value would be equal (C_P/C_V) ratio.

This error in ‘k’ also illustrates the importance of specifying which correlation is to be used when ordering a performance test (ie, refer to ASME PTC-10 for additional details), so that client and vendor are on the same agreement moving forwards with regard to molecular weight (MW) and k for the test fluid. For further detail refer to reference [6] and August and September 2010 TOTMs [7, 8].

It may also be worth noting that when trending ‘n’ and the polytropic efficiency to evaluate machine condition, the relative accuracy of measurement instrument/equipment (temperature and pressure transducers) and mapping of compressor performance to the original performance curve (actual gas volume flow rate vs speed), introduces many potential erroneous sources into this daily evaluation.

Note that the accuracy of the shortcut methods is dependent on the values of k and n. The definition of average temperature in the shortcut method was adjusted to obtain a better match between the isentropic path exponent (k) calculated by rigorous method.

To learn more about similar cases and how to minimize operational problems, we suggest attending our G4 (Gas Conditioning and Processing), PF4 (Oil Production and Processing Facilities), ME46 (Compressor Systems–Mechanical Design and Specifications) and ME44 (Fundamentals of Pump and Compressors Systems), courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

Dr. Mahmood Moshfeghian

Reference:

1. Maddox, R. N. and L. L. Lilly, “Gas conditioning and processing, Volume 3: Advanced Techniques and Applications,” John M. Campbell and Company, 2^nd Ed., Norman, Oklahoma, USA, 1990.

2. Campbell, J.M., Gas Conditioning and Processing, Volume 2: The Equipment Modules, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.

3. Soave, G., Chem. Eng. Sci., Vol. 27, pp. 1197-1203, 1972.

4. Peng, D. Y., and Robinson, D. B., Ind. Eng. Chem. Fundam., Vol. 15, p. 59, 1976.

5. ProMax 3.2, Bryan Research and Engineering, Inc, Bryan, Texas, 2014.

6. ASME PTC-10, “Performance test Code on Compressors and Exhausters”, 1997.

7. Honeywell, J. “Important Aspects of Centrifugal Compressor Testing-Part 1”, Tip of the Month, August 2010

8. Honeywell, J. “Important Aspects of Centrifugal Compressor Testing-Part 2”, Tip of the Month, September 2010

For transportation of crude oil, the pumping power requirement varies as the crude oil viscosity changes. Increasing °API or line average temperature reduces the crude oil viscosity. The reduction of viscosity results in higher Reynolds number, lower friction factor and in effect, lower pumping power requirements.

In the March 2009 tip of the month (TOTM), procedures for calculation of friction losses in oil and gas pipelines were presented. The sensitivity of friction pressure drop with the wall roughness factor was also demonstrated. In the August 2009 TOTM, we also demonstrated the effect crude oil °API and the pipeline average temperature on the pumping requirement.

In practical situations, an originating station takes crude out of storage and the midline stations taking suction from the upstream section of pipeline. Oil in the tank is often at ambient temperature, whereas once in the pipeline, the oil cools (or warms) to the same temperature as the ground.  In some parts of the world, the tank might be at +38 °C (+100 °F).  The first midline pumping station could operate at 18 °C (65 °F), and all subsequent pumping stations might operate at ground temperature, or notionally 9 °C (48 °F) with some seasonal variation. Therefore, a sound pipeline design should consider expected variation in crude oil viscosity which is normally a function of crude oil °API, and the line average temperature. In addition to °API and temperature, chemical additives may also affect crude oil pipeline pressure drop.

To reduce pressure drop and increase pipeline capacity, oil industry has utilized drag reducing agents. Drag-reducing agents, or drag-reducing polymers, are additives in pipelines that reduce turbulence in a pipe. Usually used in petroleum pipelines, they increase the pipeline capacity by reducing turbulence and therefore allowing the oil to flow more efficiently [1]. In addition to drag reducing agents, another group of chemicals called “Incorporative Additives”, which reduces crude oil viscosity, may be used. Halloran presented a series of general reading articles on chemical additives [2-4].

In this TOTM, we will demonstrate the effect of an incorporative additive on crude oil viscosity and consequently on pressure drop for crude oil pipeline transportation.

Case Study: Part 1 – Viscosity Reduction

The laboratory measured kinematic viscosity for different °API crude oil samples without and with “Incorporative Additive” at 50 °C (122 °F) reported by Oil Flux Americas [6] are shown in Table 1. The calculated density, absolute viscosity and percent reduction in viscosity for each oil sample at 50 °C (122 °F) are also shown in this table. As noted in this table, the lower °API (heavier oil), the greater the reduction in oil viscosity. The measured kinematic viscosities as a function of crude oil °API are shown in Figure 1. The absolute viscosity is calculated by multiplying the measured kinematic viscosity by density. The corresponding calculated absolute viscosities are also shown in Table 1 for crude oil samples “Without” and “With” additive, respectively.

Table 1. Measured kinematic viscosity [6] and absolute viscosity for several crude oil samples at 50 °C (122 °F) without and with chemical additive.

Molecular sieves are used upstream of turboexpander units and LNG facilities to dehydrate natural gas to <0.1 ppmv water content.   In the natural gas industry, the molecular sieves employ heat to drive off the adsorbed water.  Figure 1 shows a typical flow schematic for a 2 tower system; Figure 2 shows a 3 tower system.

The cyclical heating/cooling of the adsorbent results in a capacity (kg water/100 kg adsorbent; lb_m water/100 lb_m adsorbent) decline due to a gradual loss of crystalline structure and/or pore closure.  A more troublesome cause of capacity decline is contamination of the molecular sieves due to liquid carryover from the upstream separation equipment.

Figure 3 shows a generic molecular sieve capacity decline curve.  A few important observations can be made from Figure 3:

1. The life of the adsorbent is a function of the number of cycles, not the elapsed calendar time.
2. The capacity decline is steep at the beginning but gradually flattens out. This assumes no step-change events such as NGL, glycol, and/or liquid amine carryover, bed support failure, etc.
3. Shown in this figure are “Good”, “Average” and “Poor” curves that are a function of site specific factors.
4. Locating one data point on Figure 3 from a performance test allows you to extrapolate the decline curve of the unit in question.

If your regeneration circuit has excess capacity over the “normal design conditions”, i.e., a design factor, you have standby time.   This excess capacity allows you to reduce your online adsorption time and “turn the beds around” faster by regenerating the beds in a shorter cycle time.   When you are involved in the design of an adsorption unit, it is recommended to add 10 – 20% excess regeneration capacity.

Because of the capacity decline curves flatten out, available standby time may be able to extend the life of a molecular sieve unit when your unit is operating on fixed cycle times.  Other operating options include: running each cycle to water breakthrough; and, reducing the cycle times in discreet steps throughout the life of the adsorbent.

 

To illustrate the benefits of standby time, consider the following case study.   A natural gas processing plant has commissioned a new 3 tower molecular sieve dehydration unit to process 11.3 x 10⁶ std m³/d (400 MMscfd) prior to flowing to a deep ethane recovery unit.  The unit is expected to run for 3 years before needing a recharge and the plant turnaround is based on this expectation.  The following assumptions are made:

• 3 tower system (2 towers on adsorption, 1 on regeneration)
• External Insulation
• Tower ID = 2.9 m (9.5ft)
• Each tower contains 24630 kg [54300 lbm] of Type 4A 4×8 mesh beads
• Regeneration circuit capable of handling an extra 15% of flow
• Unit is operated on fixed time cycles
• No step-change events such as liquid carryover, poor flow distribution, etc.

The design basis and molecular sieve design summary are shown in Tables 1 and 2.  The additional 15% of flow from the regeneration gas heater is well below the point at which bed lifting will occur.

The calculations presented here are valid for low pressure regeneration (less than 4100 kPaa (600 psia).  Using the concepts outlined in Chapter 18 of Gas Conditioning and Processing, Volume 2 [1]:  The Equipment Modules (9^th Edition) we find a design life factor, F_L, of 0.6 after 3 years (1 095 cycles) of operation at design conditions.  This point lies slightly above the “average” life curve as seen in Figure 4.

After 12 months of operation, a Performance Test Run (PTR) is conducted.  The results are shown in Table 3.  The feed flow rate and temperature are slightly lower compared to the design values.  A water breakthrough time of 20.9 hours is recorded.  The F_L is determined (using the concepts in Chapter 18) to be 0.68 after 365 cycles (one year of operation).  It is important and useful to understand the equation sequence of the concepts in Chapter 18, as shown by Equations 18.5 through 18.10 to arrive at the cited value for FL.   This data point is shown in Figure 5 and is seen to lie just below the generic “Average” curve.   Note that the slope of the curves are starting to flatten out.  Since the PTR F_L is lower than the Design F_L, the molecular sieves will experience water breakthrough if operated at design conditions in less than three years. Figure 6 shows the projected life factor, F_L, after 3 years of service at design conditions.  If the capacity decline continues to follow the same trend as seen from the PTR, water breakthrough will occur after 750 cycles or just a little over 2 years from startup if operation continues at design conditions.  This is shown in Figure 7.

Table 3. Results of Performance Test Run (PTR) after 12 months of operation

            Because the unit has a regeneration circuit that can handle an additional 15% of flow, the complete regeneration cycle (heating, cooling, de- and re- pressurization) can be reduced to 7.0 hours.  This allows the beds to turn around faster.

Using the reduced cycle time (the complete cycle time is now 21 hours vs the original 24 hours), we find an F_L = 0.53.  This is because less water is being adsorbed per cycle.  This occurs at around the 1500 cycle mark as shown in Figure 8.

If the plant elects to take advantage of the standby time and operate at reduced cycle time immediately following the PTR, the molecular sieves should last an additional 2.7 years, resulting in a total life of 3.7 years.   In this case, standby time will allow the unit to operate until the scheduled plant turnaround.

            We can draw the following conclusions from this case study:

1. The methods presented allow the user to estimate the decline of their adsorbent based on only one performance test run for molecular sieve dehydrators using low pressure regeneration. This permits early formulation of a credible action plan.
2. Site-specific factors will determine your unit’s decline curve. Consequently, conducting more than one performance test is highly recommended.  A poorly performing inlet separator, for example, could result in a unit exhibiting a more pronounced decline than indicated by the generic curves in Figure 3.
3. Standby time offers a large degree of operating flexibility because the decline curves tend to level off; always try to build in standby time in any new molecular sieve design.
4. Adsorption capacity is a function of the number of cycles, not calendar time.
5. Install a good filter coalescer or filter separator upstream of your adsorption unit to keep the contaminants out of the system.

The approach discussed in this Tip of the Month should help a facility engineer plan for the inevitable replacement of molecular sieves in their natural gas dehydration facility.

To learn more about similar cases and how to minimize operational problems, we suggest attending our G4 (Gas Conditioning and Processing) and PF4 (Oil Production and Processing Facilities) courses.

PetroSkills offers consulting expertise on this subject and many others. For more information about these services, visit our website at http://petroskills.com/consulting, or email us at consulting@PetroSkills.com.

By: Harvey M. Malino

Reference:

1. Campbell, J.M., Gas Conditioning and Processing, Volume 2: The Equipment Modules, 9^th Edition, 2^nd Printing, Editors Hubbard, R. and Snow–McGregor, K., Campbell Petroleum Series, Norman, Oklahoma, 2014.