Hydrocarbon Processing Copying and distributing are prohibited without permission of the publisher
Email a friend
  • Please enter a maximum of 5 recipients. Use ; to separate more than one email address.

Improve performance of exchangers in quench-water services

06.01.2014  |  Lang, T.,  Wieland-Werke AG, Ulm, GermanyProvost, J.,  Technip , Paris, FranceEl-Hajal, J. ,  Wieland-Werke AG , Ulm, Germany

Keywords: [heat exchanger] [boilers] [tubes] [ethylene] [olefins] [condenser] [alloys]

Dual-enhanced (DE) tubes are widely used in shell-and-tube heat exchangers (HEXs) in a range of industries. They have become a standard in multiple clean services for the hydrocarbon processing industry (HPI) over the last 15 years. Yet, operators as well as manufacturers struggle with the lack of operating data for DE tubes, especially in fouling type services. This article reports a successful field test of DE tubes in a quench-water service for a naphtha-based olefin plant.


Low-finned (LF) and DE tubes are widely accepted as standard solutions in shell-and-tube HEXs in many industries such as air conditioning and refrigeration, power, the HPI and machinery and equipment industry. These applications are considered clean, and the process conditions are controlled. If necessary, adequate protection, mitigation and cleaning strategies are considered.

For the HPI, substantial progress has been made in identifying the qualifications for enhanced heat-transfer technologies in clean services such as key reboiler and condenser HEXs both in process refrigeration and thermal distillation processes, e.g., in natural gas liquefaction and ethylene plants. However, substantial uncertainty exists in broader applications, especially under fouling applications.

Besides laboratory testing, field qualification for any industrial application is needed. A recent field trial with quench water studied the fouling behavior of DE tubes.a A tube bundle was installed as a reboiler in a C3 stripper for an ethylene plant during a plant shutdown. Quench water provided heating on the tube side of the reboiler.

In many cases, the limiting factors in using LF and DE tubes are attributed to the lack of information, especially in fouling situations. In the HPI, operating conditions can vary substantially. But, process HEXs must perform efficiently for extended periods, at least between shutdown cycles. Additional cleaning of the HEXs contributes to unwanted downtime, lost production and, in many cases, complex procedures.

Cooling water systems. DE tubes in copper and copper-nickel are widely used in cooling water systems with an open cooling tower in the air-conditioning and refrigeration industries. In these applications, proper functioning of anti-scaling, particulate fouling and biocides programs are critical. Extensive studies have confirmed stable operation of distinct DE tubes with sophisticated internal fin structures.2–4

HPI services. In the HPI, LF tubes have been applied successfully in fouling services such as refinery reboiler and condensers, along with crude oil pre-heating trains.5–7 Most of these references are linked to debottlenecking and capacity expansion projects. Additional energy consumption, as well as all related operational costs due to fouling, are estimated to be 0.25 % of the gross domestic product of the industrialized countries; the attractiveness to improve the operability of HEXs and process units is very high.8

Over the last 15 years, DE tubes in carbon steel (CS) have been applied and used as standard technology in liquefied natural gas (LNG) and ethylene plants.1, b The primary applications are for clean refrigeration and separation processes with clean C2 and C3 hydrocarbons and refrigeration fluids.

To further broaden the application of enhanced heat transfer technologies, manufacturers are working on using DE tubes for HEXs handling cooling water as well as quench water in ethylene plants. For cooling water applications, existing experience and design information in the air-conditioning and refrigeration industry can be transferred to designing condenser or gas- and liquid-cooling applications in the HPI.

In the case of quench water, further testing is required to qualify DE tubes for reboiler applications. To conduct the field test, a European operator allowed in one of the naphtha crackers to replace an existing plain-tube, stab-in reboiler of a C3 stripping column with a new HEX equipped with the DE tubes.c The requirement for the field test was proper instrumentation with flowmeter, temperature and pressure sensors.

The thermal rating of the stab-in reboiler has been carried through by an ethylene technology licensor, as shown in
Fig. 1.d The replacement heat exchanger was fabricated by a local vendor.

  Fig. 1.  Plant layout of naphtha cracker.


The purpose of a naphtha steam cracker is to produce light olefins such as ethylene, propylene and butadiene. These compounds are used in the production of polymers, rubber and other applications.

Ethylene plant

Naphtha, diluted with steam, is cracked in the cracking furnaces at temperatures of 800°C to 850°C. The resulting mixture of steam, hydrogen and hydrocarbons is called cracked gas. This gas is cooled by several steps before being compressed in the cracked-gas compressor.

One of the cooling steps includes direct contact with cold quench water, which is fed to the top of the quench-water column. The hot quench water leaves the column via the quench-water settler at 80°C. To improve energy efficiency, the warm quench water is used in downstream unit operations to recover heat. Surplus energy from the quench water removed by cooling water and air coolers, which is then lost to the atmosphere.

The cracked gas is separated via cryogenic operations and distilled into different fractions, one being the C3 fraction at the top of the depropanizer column. The C3 fraction contains propylene (C3H6), propane (C3H8), methyl-acetylene and propadiene (C3H4, named MAPD). MAPD is hydrogenated to propylene and propane with hydrogen coming from the cryogenic section in the MAPD hydrogenation reactor. This hydrogen stream also contains methane due to the cryogenic step.

The purpose of the C3 stripper is to remove light components, as well as the remaining C2 cut that may come from external C3 sources such as refinery-grade propylene.

The C3 stripper is reboiled by quench water and condensed against propylene refrigerant. The reboiler is always fully submerged in the boiling C3-liquid, and the boil-up ratio is controlled by the quench-water flowrate. The C3 stream from a naphtha cracker is usually 95 wt% propylene with 5% propane and few C4s, which are separated in the downstream C3 splitters.

Several process parameters are monitored. The flowrate of the quench water and the sump draw-off are measured by flow orifices. The inlet and outlet quench-water temperatures are measured by the PT temperature sensor. Also, the saturation temperature of the boiling C3 is measured at the sump draw-off by a PT temperature sensor. The pressure differential in the quench-water system is measured by membrane pressure transmitters.

Quench water

In steam crackers, ethylene is produced by thermal cracking in furnaces. The cracked gas is rapidly cooled to prevent product degradation and poor product yields due to secondary cracking reaction. The final cooling (175°C to 40°C) before gas compression is achieved by the two towers in series—the primary fractionator followed by a quench-water tower. The cracked gas is cooled in the quench tower by direct contact with a closed-loop, circulating quench water in the tower. The warm quench water is later used as a heating medium, e.g., for C3 splitting applications such as the C3 stripper-reboiler.

When evaluating the composition of the quench water, little accurate data are available. Unfortunately, the best answer given is: “Quench water is quench water.” The composition of quench water will vary depending on several process parameters, including:

  • Cracker type—Gas vs. naphtha
  • Available water quality
  • Management of quench-water cleanup (draining, filtration, etc.)

Tables 1 and 2 summarize data collected on several projects for both gas- and naphtha-cracking processes. In the case history, a composition analysis was done on the circulating quench water from the naphtha cracker operated by the Institut Alpha, as shown in Table 3.9



The KW-Index covers components of mineral oil between C10 and C40 with a boiling point between 175°C and 525°C.10 Volatile aromatics are primarily benzene, toluene, ethylbenzene and xylene, also collectively called BTX. Short-chain hydrocarbons are not covered by the KW-Index, and they can act as solvents. The polycyclic aromatics hydrocarbons cover a very wide group of different components. Table 3 lists the light polycyclic aromatics. No heavier components indicating the presence of tar are measured.


Fig. 2
represents typical quench-water fouling in a gas cracker plant. The industry standard of fouling management is to promote a high flow velocity to increase the shear stress at the tube wall, thereby minimizing deposits.

 Fig. 2. Tar deposits on a plate of a plate-and-
 frame HEX using quench water.


The reboiler installed at the bottom of the C3 stripping column, as shown in Fig. 3, is performing partial vaporization of a mixture made of 95 wt% propylene and 5 wt% propane with traces of ethane and lighter products. The reboiler operates at 28°C and 13.5 bara. The energy to the column is provided by quench water at 78°C and 4.5 bara. The design duty is 649 kW. This exchanger operates in pool boiling mode. Commercially available software was used in the design.11, e

  Fig. 3. C3 stripping unit with stab-in HEX
  and instrumentation.

New DE tube design criteria

When designing the new DE tube HEX, several factors were considered:

  1. Providing a design fully compliant with codes, standards and operator specifications
  2. Providing a HEX perfectly suitable as a replacement-option without any changes to surrounding piping and connections.
  3. Providing the operator with a HEX that will perform the required duty and meet the allowable pressure drop. Furthermore, no disturbance should occur during plant operations.
  4. Developing a robust HEX using adapted DE tubes that will efficiently perform with quench water.

To address first and second items, designers elected to keep the same shell diameter, connection size and location. They worked with a HEX manufacturer as recommended by the operator. Regarding the third and fourth items, the heat-transfer area was defined considering the flow regime inside the tubes and the risk of fouling.

HEX design comparison

The existing HEX is a B-U TEMA type, i.e., a B-type head for quench-water distribution with U tubes. Table 4 is the design comparison of the tube bundle for both the plain and DE tubes. With the four-pass design, the quench-water velocity inside the tube is of 0.31 m/s. The recommended water velocity should be in the range of 1 m/s to 2.5 m/s to ensure sufficient heat transfer and to limit fouling.d Since the operator allowed for a higher pressure drop, the four-pass tube-side design was replaced with a six-pass design. The new design with DE tubes yielded a water velocity of 1.5 m/s.c


For the HEX with plain tubes, the inside and outside fouling thermal resistances are applied to the bare tube areas, respectively. Under fouling conditions, the total fouling thermal resistance represents over 40% of the total thermal resistance.

Since the C3 mixture is a clean product, no fouling is considered for the external enhanced area. Nevertheless, quench water is a fouling fluid; so, internal fouling resistance was kept and applied to the internal enhanced surface. However, with the enhanced surfaces on both sides, the overall heat transfer increased substantially, thereby increasing the contribution of the fouling resistance to the overall heat-transfer resistance.

This result of the thermal design spurred motivation for a field test. In comparative quench-water services of reboilers in C3 separation sections in olefin plants, the typical fouling resistance for the quench water is 0.00034 m2K/W to 0.0006 m2K/W, as summarized in Table 5. Such large resistances dominate the design of these HEXs and results in both large heat transfer surfaces as well as an increased number of shells per unit.


The tube pitch was also increased to homogeneously distribute the tube arrangement on the tubesheet and to manage the increased vapor rate from the high effectiveness of the tubes, as listed in Fig. 4, for tubesheet layout comparison.

  Fig. 4. Tube layout of plain (a) and 
  replacement tube bundle (b).


While in operation, data were recorded every two hours for approximately 11 months (October 2012 to September 2013):

Tube side:

  • Quench water inlet temperature, Twi
  • Quench water outlet temperature, Two
  • Quench water mass flowrate, Mw
  • Quench water pressure drop.

Shell side:

  • The saturation temperature of the hydrocarbon mixture.

Fouling resistance

Two methods have been used to determine the fouling heat-transfer resistance and are described here. The measurement results from Fig. 5 show both the saturation temperature of the hydrocarbon mixture and the quench-water flowrate. Fluctuations are due to the change in service conditions for this unit. The possibility of relating fouling due to pressure drop of the quench water was investigated. Unfortunately, this information could not be used since the pressure sensor is located upstream of the flow-control valve. It resulted in a combined value of the pressure drop at the valve and from the tubes.

  Fig. 5. Quench-water flowrate and mixture
  saturation temperature.

Data reduction method

The duty of the HEX can be determined based on the inlet/outlet temperature difference and mass flowrate of the quench water as:

Q = Mw × Cp × (Twi – Two )      (1)

The duty can also be expressed as:

Q = Uo × Ao × LMTD      (2)

The overall heat transfer coefficient, Uo, corresponds to:



and the wall thermal resistance is:




First method

Assume that there is no fouling on the shell side. In this case, Eq. 3 can be expressed as:



The overall heat transfer coefficient, Uo, can be calculated from Eq. 2. The thermal wall resistance, Rw, can be determined from the tube geometry and thermal conductivity of the tube material. Further, the heat transfer coefficients on the shell side, ho, and on the tube side, hi, are obtained from laboratory measurements with water and pure propane. The weak point of this method is related to the slight difference between laboratory and plant process-fluid qualities.

Second method

Here, it is assumed that the HEX is clean at startup. In this case, the overall heat transfer coefficient, Uclean , yields:



When fouling begins, the overall heat transfer coefficient, Uf, is expressed as:



Only the experimental data for which the saturation temperature is between 23°C and 28°C is considered. Fig. 5 shows that only a few points at startup (32 over 1,355 points) were not in this range. Based on the actual heat flux analysis and laboratory measurements of the heat transfer coefficient shell side, it can be considered that the heat transfer coefficient shell side is constant for all measured data. Also, assume that there is no fouling on the shell side.

When the tube-side heat transfer coefficient, hi, is the same for the clean and the fouling cases, i.e., the same velocity, Eq. 6 is subtracted from Eq. 7 to yield:



Consequently, the tube-side fouling factor is:



At this stage, the tube-side fouling resistance is determined for the clean overall heat transfer coefficient, Uclean. Fig. 6 shows the overall heat transfer coefficient vs. Reynolds (Re) number. The data points (blue dots) were taken during the first 190 hours after plant startup and are considered to be clean conditions. The data after 190 hours of operation are shown (green squares).

For the first 190 hours, there was an excellent agreement between the overall heat transfer coefficient and the Re number, as shown in Fig. 6. For a given Re number, this linear fit will give the overall heat transfer coefficient considered under clean conditions, Uclean.

  Fig. 6. Overall heat transfer coefficient vs. Re.

After 190 hours, the data points are not in good agreement. For a given Re number, i.e., given velocity, the overall heat transfer coefficient, Uf, decreases due to the fouling thermal resistance. In combination with Eq. 9, the information facilitates estimating the tube-side fouling thermal resistance. The weak points of this method are the assumption of constant shell-side heat transfer coefficient and that no fouling occurs in the early hours of operation.

Field test results

There is some uncertainty in the measurements since the instruments and their corresponding accuracy were selected to monitor the process and not to measure the fouling thermal resistance. The quench-water inlet and outlet temperatures are not measured directly at the inlet and outlet of the HEX but a few meters away, as illustrated in Fig. 3. In Fig. 7, the fouling thermal resistance is described by the two methods described previously.

  Fig. 7. Fouling thermal resistance tube side.

The fouling thermal resistance calculated with the first or the second methods has the same behavior with some shifting in the absolute value. The rapid changes in the fouling thermal resistance values over time stemming from transient operating conditions.

Negative fouling thermal resistance, which is not possible, is related to the way the reference line with a linear regression is calculated. Even under fouling conditions, the fouling thermal resistance is smaller than 0.00015 m²K/W. The ethylene operator is completely satisfied by the performance of new equipment.


Field-test data on quench water behavior was collected to improve the knowledge regarding the behavior of DE tubes in such services. Today, the full production of an enhanced HEXs is possible due to more than 18 months of operations experience to confirm the suitability of DE tubes. It has been shown that the fouling level for the quench-water side can be reduced by a factor of three, and thereby, allowing for a substantial design and efficiency improvement in such “fouling services.” HP


Ao  Outside envelope surface at fin tip, m² 
Ai  Internal surface at internal root fins, m² 

Quench water heat capacity, J/kgK


Tube outside diameter, mm


Tube inside diameter, mm


Heat transfer coefficient shell side, W/m²K


Heat transfer coefficient tube side, W/m²K


Wall thermal conductivity, W/mK


 Log-mean temperature difference 

Mw   Quench water mass flowrate, kg/s

Wall thermal resistance, m²K/W

Rfi Fouling factor tube side, m²K/W

Fouling factor shell side, m²K/W


Duty, W


Quench-water inlet temperature, °C


Quench-water outlet temperature, °C


Overall heat transfer coefficient, W/m2K


Overall heat transfer coefficient, clean condition, W/m2K


Overall heat transfer coefficient, fouling condition, W/m²K


Hydrocarbon molecule with 2 C atoms


Hydrocarbon molecule with 3 C atoms 


The authors thank the operator of a naphtha cracker for the acceptance of these field trials. Due to company policy, the participation of the partner could not be mentioned officially. The authors appreciate and acknowledge the excellent support throughout the whole project.


a Technip and Wieland tested the DE tubes in industrial conditions, thanks to a well-known European olefin plant operator.
b Wieland’s enhanced GEWA-PB and –KS tubes in CS have been applied in LNG and ethylene plants.1
c DE enhanced GEWA-PB tubes
d Technip is the ethylene technology licensor.
e HTRI software was used for the design using correlations for the GEWA-PB tube developed by Technip France Heat Transfer Department and Wieland-Werke AG.11


1 Ploix, B. and T. Lang, “Application of dual enhanced Wieland tubes for ethylene and LNG services,” 2010 AIChE Spring National Meeting, San Antonio, March 21–25, 2010.
2 ANSI/AHRI Standard 550/590 (I-P), Standard for performance rating of water-chilling and heat pump water-heating packages using the vapor compression cycle, Air-Conditioning, Heating and Refrigeration Institute, 2011, www.ahrinet.org.
3 Webb, R. L. and N.-H. Kim, “Particulate fouling on enhanced tubes,” National Heat Transfer Conference 1989, Philadelphia, August 1988.
4 Hays, G., E. S. Beardwood and S. J. Colby, “Enhanced heat exchanger tubes: Their fouling tendency and potential cleanup,” ECI Symposium Series, Vol. RP2, 6th Int.l Conf. on Heat Exchanger Fouling and Cleaning, Kloster Israel, June 5–10, 2005.
5 Webber, W. O., “Does fouling rule out using low fin tubes in reboilers,” Petroleum Refiner, Vol. 39, pp. 183–186, Gulf Publishing Company, 1960.
6 Moore, J. A., “Fintubes foil fouling for scaling services,” Chemical Processing, August 1974.
7 Miller, E. R., “Performance of low fin tubing in asphalt service,” Heat Transfer Proc. Int. heat Transfer Conf., 7th, Vol. 6, pp. 409–414, 1982.
8 Müller-Steinhagen, H., M. R. Malayeri and A. P Watkinson, “Recent Advances in Heat Exchanger Fouling research, Mitigation, and Cleaning Technologies,” Heat Transfer Engineering, Vol. 28, No. 3, pp. 173–178, 2007.
9 Institut Alpha GmbH & Co. KG, Dornstadter Weg 15, 89081 Ulm -Jungingen, Germany.
10 Heat Transfer Research, Inc.
11 ISO 9377-2:2000 Water quality, Determination of hydrocarbon oil index, Part 2: Method using solvent extraction and gas chromatography.

The authors 
Jeremy Provost is the manager of the heat transfer department—Process and Technology Division, Technip, Paris, France. He has over 16 years of experience in heat exchanger design for O&G applications. He started as heat transfer engineer with GEA BTT and was a leader in the air-cooled heat exchangers for O&G. He later became the manager of the heat transfer department. Mr. Provost was involved in several high profile LNG projects. He joined Technip France heat transfer department. Mr. Provost is member of the HTRI communication committee. He holds a diploma in heat transfer engineering from the Engineering School of Marseille and a diploma in fluid dynamics at the Marseille- Provence University, France. 
  Jean El Hajal is senior R&D heat transfer engineering at Wieland-Werke AG. He has over 10 years of experience in the development of enhanced heat transfer surface for one phase flow, evaporation and condensation. He also has extended experience in CFD simulation and heat transfer software. He received his PhD in mechanical engineering from the University of Nantes, France. Dr. El Hajal worked as postdoctoral researcher at the LTCM laboratory at the Swiss Federal Institute of Technology (EPFL). He has published several papers on evaporation and condensation heat transfer. 
Thomas Lang is manager technical marketing for process industry of Wieland Thermal Solutions, at Wieland-Werke AG, Ulm, Germany. He has over 20 years of experience in enhanced heat transfer for the hydrocarbon processing industry. He also has experience in the power, air-conditioning and refrigeration industries. He is member of the HTRI communication committee. Mr. Lange received his BS degree in mechanical engineering at the University of Stuttgart, Germany and an MS degree in civil engineering from the University of Colorado. 

Have your say
  • All comments are subject to editorial review.
    All fields are compulsory.

Related articles


Sign-up for the Free Daily HP Enewsletter!

Boxscore Database

A searchable database of project activity in the global hydrocarbon processing industry


Is 2016 the peak for US gasoline demand?




View previous results

Popular Searches

Please read our Term and Conditions and Privacy Policy before using the site. All material subject to strictly enforced copyright laws.
© 2016 Hydrocarbon Processing. © 2016 Gulf Publishing Company.