A. Sharma, Contributing Author, Jaipur, India
Of the condensers in process industries such as refineries and petrochemical facilities, many are flooded-type condensers. Flooded condensers are fundamental heat exchangers that condense hot vapor or gas into a liquid. During typical operation, 10%–40% of the condenser's area is covered or filled with liquid, while the balance is used for vapor condensation.
These flooded condensers are typically employed in total condensers, where hot vapor fully condenses into liquid and the outflow is completely liquid. Usually, pure liquid is produced by fully condensing hot vapor. This means that ethylene and propylene fractionators, debutanizers, depropanizers, depentanizers, iso-butane and propylene compressors and other systems with mostly pure hot vapor can be developed with total and flooding condensers.
Flooded condensers are generally favored in pure component condensers since their dew point and boiling point (bubble point) are the same for pure (single component) components. A small pressure drop can drastically vaporize the condensed liquid in the reflux drum and lower the liquid level. Subcooling of the liquid is essential to avoid such a level fluctuation; therefore, flooded condensers are chosen to provide subcooling.
Typical flooded condenser designs fall into two categories. The first contains a control valve installed at the condenser inlet or outlet and the condenser outlet is submerged in the reflux drum. Due to backpressure, the condensed liquid is resisted and flows back into the condenser (FIG. 1).
In the second design, the reflux drum is elevated above the condenser and liquid is lifted to the reflux drum due to liquid subcooling (FIG. 2). This article will delve into following challenges and providing recommendations for condenser design to avoid process disturbances.
Challenges. Non-condensable gas buildup primarily affects this flooded condenser. Small bubbles can clump together and generate vapor pockets. Non-condensable gases such as methane and hydrogen (H2), and inert gases such as nitrogen, air and carbon dioxide (CO2) build in the flooded condenser because its bottom section or condenser outlet is back flooded with liquid, which acts as a liquid seal for this vapor and prevents it from escaping through the liquid. Vapor tends to lift rather than fall due to gravity.
This inert or non-condensable gas insulates by covering the tubes and minimizes heat transfer (the vapor heat transfer coefficient is very low), known as vapor binding. In 1983, Bell, et al., stated that more than half of all condenser problems are due to poor venting.1
One of the fluid catalytic cracker (FCC) facilities' debutanizer towers had a flooded condenser (Type 1), and the reboiler was swapped over to stand by after nitrogen pressurization-depressurization (PDP). This residual nitrogen within the column accumulated in the top condenser as overhead vapor. The nitrogen concentration was high enough to destabilize the column. Column pressure varies, and the column tripped due to high pressure. The issue was that there was no vent nozzle on the flooded condenser's shell side to vent out non-condensable gases.
Another challenge comes from an ethylene plant, where the pressure in the ethylene fractionator (C2 splitter) varies at high pressure despite a high propylene flow (as a coolant) in the condenser. The non-condensable vent nozzle was present to vent out; however, it was blinded without an isolation valve.
The iso-butane refrigeration compressor in the refinery tripped on high pressure due to an accumulation of nitrogen in the system—the flooded condenser is the only place where the nitrogen can vent out inert or non-condensable (lighter) gas to minimize hydrocarbon loss. The vent was present, but it took a long time to vent non-condensable gas because the vertical baffle in the heat exchanger trapped most of the non-condensable gas and impeded the vapor flow, limiting the heat transfer area (FIG. 3).
A similar issue occurred in a steam cracker where the propylene refrigeration compressor faced a high-pressure scenario: the vent location was incorrect, resulting in the high likelihood of inert accumulation. Additionally, numerous case studies have also been reported.2,3
The preceding situation demonstrates why a permanent vent with an isolation valve is essential in a flooded condenser and why the vent should be located where there is a high risk of inert accumulation.
Solutions. These flooded condensers necessitate the installation of a permanent vent with a valve to exhaust non-condensable gases. It is always preferable to redirect into a flare so that non-condensable gases can exit the system. Some plant vents are, in some manner, linked to reflux drums. The gases have been vented but are still in the system. When hot vapor condenses in the shell side of the condenser, the vent must be located far enough away from the vapor inlet to avoid bypassing hydrocarbon vapor rather than inerts.
A vertical baffle is often installed on the shell side of the condenser to support the tube and produce turbulence in the fluid, improving heat transmission. These baffles also trap most non-condensable gas, preventing it from reaching the vent, always resisting heat transmission and bottlenecking the condenser.4 Drilling a 6-mm hole at both the top and bottom is always preferable to vent non-condensable gas and drain-trapped liquid (FIG. 4).
If the vapor condenses in the tube side of the flooded condenser, most non-condensable gases are prone to become trapped in the condenser's bottom pass partition of tube bundle; however, vent nozzles are often provided in the condenser's top compartment. Venting should always be done through the bottom pass partition of the tube compartment.5
When the bottom section of the partition vent is unavailable, an alternative method is to drill a 6-mm hole in the pass partition plate to let non-condensable gas enter the top compartment and exit through the vent to flare (FIG. 5).
The flood-back varies in flooded condensers; therefore, it must be considered. Due to backpressure in the reflux drum caused by friction drop, piping or other drum restrictions can flood the condenser. The safety factor in the surface area of the condenser must be greater than the normal necessary for heat transfer for condensation.6,7
A steam reboiler can have similar flood-back and variable area issues. Where the condensate header is well elevated, and the condensate cannot lift, the steam reboiler is back-flooded with water, reducing the heat transfer area and bottlenecking the reboiler.
Thermal scanning is the best way to diagnose the problem of inert accumulation and liquid level in the condenser. A thermal assessment of the shell can provide enough information regarding its current state. The age of the condenser is also affected by inert buildup. Heat transfer is reduced because the flooded area of the tube is colder than the top or inert accumulated tubes, and this temperature difference affects the condenser's aging.8
To vent out non-condensable gas from flooded condensers incessantly, direct control of the flow controller is utilized in some plants where non-condensable gas slippage from upstream units (such as methane or H2) is expected in normal operation—this flow controller should be a pressure controller. This flow controller always maintains a constant amount of non-condensable flow, but some regular hydrocarbon or product is also vented out and some lighter hydrocarbons remain in the system, causing pressure to fluctuate. To avoid such a problem, always use a pressure controller instead of a flow controller.
Takeaway. Safe practices for handling a condenser include:
NOTE
This work is based on the author’s experience and learning. It is not affiliated with any company.
LITERATURE CITED
ABHISHEK SHARMA works at an ethylene plant as a process engineer. He has more than 5 yr of experience working in a steam cracker unit. Sharma earned a BS degree in chemical engineering with honors from the National Institute of Technology Raipur in India and finished a process equipment design course at the Indian Institute of Technology Roorkee. He is an associate member of the IChemE, an active professional member of the American Institute of Chemical Engineers, and the author of eight technical articles in industry publications.