Technical Report

Testing of Horizontal Falling-Film Evaporator for Produced-Water Treatment

Harold Krenkel, Manager, Processing Technologies, Alberta Innovates-Technology Futures

Part of Alberta’s research and innovation system, Alberta Innovates-Technology Futures (AITF) is helping to build healthy, sustainable businesses in the province. Through a suite of programs and services for entrepreneurs, companies, researchers, post-secondary institutions, and investors, AITF provides technical services and funding support to facilitate the commercialization of technologies, develop new knowledge-based industry clusters, and encourage an entrepreneurial culture in Alberta. AITF has an internationally recognized reputation in developing and evaluating technologies associated with in-situ bitumen and heavy-oil recovery, with unique facilities and extensive laboratory programs.


The production of bitumen by steam-assisted-gravity-drainage (SAGD) technology requires treatment of significant amounts of produced water (PW). The treated PW needs to meet boiler feedwater specifications so that it can be recycled. The water-treatment process typically used in SAGD operations is softening, using warm or hot lime followed by ion exchange. This traditional treatment approach is a mature and well-known process that is effective for hardness and silica removal. However, this process requires significant chemical inputs and highly trained operators. It also generates several solid- and liquid-waste streams.

Evaporation is an alternate water-treatment technology that uses industrial vertical falling-film evaporators to recycle 95 to 97% of the water for reuse as boiler feedwater. These systems have the advantage of reduced waste generation, reduced chemical input, and smaller footprint. However, these systems face challenges related to scaling and corrosion, as well as high capital costs.

Horizontal falling-film evaporators are designed to address some of the challenges faced by vertical falling-film evaporators in treating water from SAGD operations. Horizontal evaporators have been used in commercial operations for seawater desalination applications. On the basis of this experience, one of the potential advantages of the horizontal units over vertical evaporators is better availability and reliability as a result of the feasibility of pulling the heat-transfer tubes out for cleaning. Further, because horizontal evaporators are prefabricated, on-site installation costs could be significantly lower.

As part of its mandate, AITF provides third-party testing services to vendors developing technologies for the oil and gas sector. In this case, it has set up a testing program for a company with considerable international experience in thermal water-treatment solutions, with the purpose of evaluating horizontal falling-film evaporators for the treatment of SAGD PW and once-through steam-generator (OTSG) blowdown (BD). A thermal-based water-treatment pilot unit has been built for laboratory-scale testing. Trials are being performed with PW and OTSG BD from several SAGD sites. For all trials, the unit was operated under essentially constant process conditions for a 1-month period. The BD—feed, distillate, and brine—process streams were sampled and characterized. Upon completion of each trial period, the unit was cleaned and restarted with a new water source (Fig. 1).

Fig. 1—Horizontal falling-film evaporator pilot unit.

Process Description

PW/OTSG BD enters the system from a feed tank (Fig. 2). 

Fig. 2—System illustration.

At first stage, the feed flows to the de-aerator, where the noncondensable gases are removed and the feed is heated. The feed line is connected to the main recirculation line, where it is blended with the circulation flow.

The recirculation line is divided into two streams:

·         Recirculation stream: Sprayed onto the shell side of the evaporator-tube bundle. The dropping brine creates falling films on the outside of the heat-transfer tubes. This film then absorbs heat, which is generated by condensation occurring on the inside of the tubes, thereby causing a portion of the brine to evaporate. The vapour is then drawn into the compressor.

·         Brine stream: Concentrated brine is discharged to the brine tank. The vapour produced on the shell side of the tube bundle is compressed and discharged into the tube side of the tube bundle, where it is condensed and discharged as distillate. This distillate is then pumped into the distillate tank.

All process chemicals are dosed directly to the recirculation line.  The chemicals in the process are mainly used for pH control and to prevent scaling and foaming.

Operating Conditions. Process conditions for the three feed types are presented in Table 1.

Unit Performance. The unit performance is monitored by a calculation of the heat-transfer coefficient (HTC). The calculation is based on the unit’s production, the heat-transfer area, and the temperature differences between the side of the shell and the side of the tube (thermodynamic driving force). The following equation is used to perform the calculation:

......................................................................................................................................... (1)

Here, U [(kcal/h)/(m2×°C)] is the HTC, F [kg/h] is the distillate flow, hfg [kcal/kg] is the enthalpy of evaporation, A [m2] is the heat-transfer surface area, and ΔT [°C] is the thermodynamic driving force.

The calculation is performed initially after the system has been cleaned. The initial calculation represents a baseline value. The HTC is calculated on a daily basis and compared with the baseline value to determine whether there has been any change in the system’s performance.

The coefficient is therefore presented in relative terms, as a percentage of the cleaned state. A coefficient of 100% signifies that the HTC is identical to the baseline value that was calculated in the beginning. Any change in the coefficient indicates a change in the unit’s predefined performance parameters. A coefficient decrease indicates faults in the unit performance, which may be caused by formation of scale on the tubes, which disrupts the evaporation process. This disruption is compensated by changing the compressor working parameters, keeping evaporation capacity constant.

The HTC for the two trials with PW is shown in Fig. 3. 

Fig. 3—HTC for Trials 1 and 2 with PW.

In both trials, the HTC dropped during the first week. The creation of the initial scale caused significant thermal resistance. This resistance is expressed by a reduction in the HTC. The drop in the HTC is a common phenomenon during the initial operation of evaporators. From the second week onward, the HTC stabilized, with the final HTC value approximately 75 to 80% of the clean value. The reduction in the HTC did not affect the throughput of the evaporator because the control system compensates for this reduction by increasing the temperature difference ΔT to generate a higher thermodynamic driving force.

For the OTSG BD run, the unit operated stably except for a foaming event that happened after 2 weeks of operation. During this period, foam appeared on the evaporator shell. The foaming led to bubbles, which increased the evaporation area and resulted in an increase in the HTC (Fig. 4). 

Fig. 4—HTC for Trial 3 with OTSG BD.

Subsequently, the foaming was brought under control, and the HTC stabilized, with the final HTC value being approximately 70 to 80% of the clean value.

Liquids Chemistry. The water analysis of the second trial with PW, shown in Table 2, includes the results of feed, brine, and distillate analyses.

·         The calculated concentration factor (CF), using potassium as a representative ion because of the high solubility of potassium salts, is 32, indicating 97% recovery. CF is calculated as the ratio of the ion concentration in brine (489 ppm) to feed (15.1 ppm).

·         Brine-to-feed ratio for magnesium and calcium is lower than the concentration factor, indicating that hardness scaling could occur. However, because the total calcium and magnesium concentrations were low, hardness scaling is negligible.

·         For silica, the same ratio is slightly lower than the CF, indicating silica scaling occurred. This was verified by solids analysis that was performed on the material recovered from the heat-transfer tubes.

·         The evaporator produced high-quality distillate. Impurities such as silica and hardness were found to be negligible. The distillate quality met the requirements for a downstream process such as drum boiler feedwater.

Tube-Bundle Cleaning. The cleaning was performed at the local Clean Harbors facility, as per the following procedure:

1.       The bundle was washed with hot water (70–80°C) to remove materials that were loosely attached (approximately 25% of the scaled materials).

2.       The tube bundle was submersed for 4 hours in a solution of 5% Paratene M390, 2% ammonium bifluoride, 2% Paratene D731, and 0.25% Paratene S620. The solution was heated to 70°C and circulated.

3.       The bundle was removed and washed again (as described in Step 1) with hot water at medium pressure to ensure full cleaning.

The cleaning was very successful, with no residual scale remaining on the tubes after cleaning. When the cleaned tube bundle was reinstalled, the evaporator-system performance returned to the same level as the first day of the trial.


A laboratory-scale testing program has been carried out to evaluate the performance of a horizontal falling-film evaporator in treating SAGD PW and OTSG BD. The results have been promising. The product—distilled water—is highly pure and, in some cases, meets the boiler-feedwater specifications, depending upon the facility requirements.  Throughout all the pilot experiments, the unit operated continuously and reliably, achieving consistent, commercially viable results.

The horizontal falling-film evaporator was designed with removable heat-transfer media (tube bundles). The capability to extract the tube bundles is a major advantage over the classic evaporator design, allowing for easy inspection and simplified cleaning. Traditionally, internal cleaning (either as part of routine cleaning cycles or following feed upsets) has required production to be halted and the system to be shut down. With a removable tube-bundle design, cleaning or tube-bundle replacement can be carried out with a short downtime and minimal impact on production.

Future testing will focus on assessment of tube-bundle cleaning techniques, as well as optimization of operating conditions, chemical selection and dosing, and scaling and fouling mitigation.