J. JOHNS, Becht, Salt Lake City, Utah (U.S.)
One of the most frustrating situations to endure is the slow cooldown of a reactor (or series of reactors) when shutting down a hydrocracker, hydrotreater or renewable feed unit. It often seems that the entire refinery is watching the reactor cool, especially as reactor temperatures drop below 200°F (93°C). This article details methods of speeding up the cooling process while maintaining equipment integrity and process safety. Investing a little time and effort into the cooldown process can pay big dividends!
Without procedures and processes to accelerate the cooldown, it can take 80 hrs–100 hrs to fully cool down a reactor. FIG. 1 shows a typical reactor cooldown curve indicating both the hottest catalyst temperature and the hottest skin temperature (typically the skin temperature on the bottom head of the reactor). Each reactor(s) may have a different cooldown time, but the curve should look similar to the one shown in FIG. 1.
Considering that maintenance costs and lost profit opportunities can range from tens of thousands to hundreds of thousands of dollars per day, the opportunity to improve reactor cooldown can easily be worth at least a few million dollars for many units.
Pull the cracked stocks early. Although the author does not fully understand the reasons, 40 yrs of experience have shown that pulling out cracked stocks 8 hrs–12 hrs before pulling all the feed helps to reduce the light hydrocarbons on the catalyst.
Maximum cooldown rate for equipment safety and reliability. A well-established maximum cooldown rate for high-pressure equipment is 100°F/hr. It is imperative not to exceed this rate to avoid high stresses in the thick metal of high-pressure equipment. During the initial cooldown, it is easy to cool down at 100°F/hr using the reactor furnace to adjust the cooling rate. The cooldown rate of 100°F/hr applies directly to the reactor skin temperatures, but this rate is also a good idea for reactor catalyst temperatures. Maintaining the same cooldown rate keeps things simple and helps keep portions of the reactor skin not covered by thermocouples within the cooling rate limits.
Smooth and steady. The cooling rate should also be smooth and steady. The author once saw an operator cool the reactor to 100°F in 15 min and then take a 45-min break! This “unsteady” cooling rate is extremely hard on equipment health.
Hot stripping: How long and how hot? The following are rough guidelines to minimize the time required for hot stripping:
An approximate rule of thumb is that the hot strip time can be cut in half for every temperature increase of 50°F (10°C), so hotter is much better.
The hot strip should be maintained for a minimum of 6 hrs–8 hrs or at least 2 hrs past the time that liquid stops appearing in the cold separator (whichever is longer).
Hot stripping should be done at normal operating pressure while maximizing the flow of recycle gas.
Reactors that have experienced maldistribution or high-pressure drop will need additional hot strip time because of the flow channels through the reactor. When in doubt, go longer and hotter to save time later.
Feed/effluent exchanger bypass. A bypass around the feed/effluent exchanger(s) (FIG. 2) decouples the heat transfer from the product to the feed and drastically speeds up reactor cooling. Any unit that does not include a feed/effluent bypass should add one soon. The bypass is also an important safety feature to help cool a reactor in an emergency. Use the feed/effluent bypass to speed up cooling when the cooling rate begins to fall below 100°F/hr.
Nickel carbonyl testing. The hazards of nickel carbonyl formation and the procedures to prevent formation above 1 parts per billion (ppb) are detailed in literature.1 As a reminder, the reactor should not be cooled down below 400°F (204°C) until the carbon monoxide (CO) content of the makeup and recycle gas is < 10 parts per million (ppm), measured with the apparatus shown in FIG. 3.
Remember that renewable units tend to produce a substantial amount of CO during normal operation, so an efficient method to remove the CO must be in place for a quick reactor cooldown.
Keep the pressure on. Maximizing recycle gas flow helps to cool the reactor faster, but more mass flow is obtained from the recycle gas when the pressure remains high as long as possible.
Brittle fracture prevention. As reactor skin temperatures approach the minimum pressurization temperature (MPT), the reactor system pressure must be lowered to prevent brittle fracture. The typical pressure limit when any metal temperature is below the MPT is typically 25%–30% of the maximum allowable working pressure (MAWP). Some units have developed MPT curves rather than an MPT point. The curves allow the pressure to stay higher while gradually reducing pressure with the change in reactor skin temperatures.
Whoa there—slow that compressor! As the reactor catalyst temperatures approach the compressor outlet temperature, slowing down the compressor will lower the temperature of the gas going to the reactor and help the reactor cool down further.
Cool nitrogen vapor injection. Many refineries bring in nitrogen pumper trucks and inject cool nitrogen vapor into the recycle stream going to the reactor. Injecting nitrogen vapor can be done without special facilities, but the temperature of the vapor going to the reactor should be maintained above 40°F (4.4°C) to minimize metal stresses in the equipment and to prevent water from freezing.
Nitrogen injection is usually started as the reactor temperatures approach ~300°F (~148.9°C) to accelerate cooling to final target temperatures. Nitrogen injection is not inexpensive, but it does help cooling and can have a good payout on the cost with time saved.
Cold nitrogen liquid injection. A few refineries inject liquid nitrogen into the recycle stream for added cooling. While liquid nitrogen injection will cool faster than cool nitrogen vapor injection, the injection point requires a specially designed quill and spool to prevent the cold liquid from cracking the steel. Metallurgists and mechanical design experts should be consulted to obtain a properly designed system for liquid nitrogen injection.
Never inject liquid nitrogen without piping specifically designed for this purpose.
Temporary cooling exchanger. A recent innovation that offers significant help in cooling a reactor is to add a temporary cooling water exchanger to the outlet of the compressor, as shown in FIG. 4. Rental companies offer exchangers that can be installed prior to a reactor shutdown. Some refineries have installed piping stubs and valves to facilitate the exchanger installation.
Water flooding the reactor. An innovation that a few companies use extensively is water flooding the reactor. Water flooding has advantages: it cools down the reactor(s) quickly and does not require inert entry into the reactor. Reactors that dump water and catalyst together will dump catalyst faster and more completely. A few disadvantages are the need to separate water from the catalyst/water slurry dumped out of the reactor and to dispose of the water.
For the water flooding process, the reactor is cooled until the hottest metal temperature is < 200°F (< 93.3°C). The water should never boil with the stainless-steel lining inside the reactor. Once the appropriate temperature is reached, the reactor is blinded and the top head removed. Water is then injected into the reactor at an appropriate, calculated rate, so as not to quench the reactor shell too quickly.
Once the reactor is filled and cooled down, the catalyst/water slurry is dumped with the reactor under an air environment. The wet catalyst prevents pyrophoric reactions. The air environment is much safer than an inert environment, and it facilitates entry into the reactor for internals inspection and catalyst loading without an inert environment.
Many years ago, a diesel hydrotreater reactor was shut down without water flooding and then, 4 yrs later, it was shut down again using a water flooding procedure. A comparison of the two reactor cooldown curves (FIG. 5) shows that water flooding cuts about one-third of the time off the total cooldown, reducing the total cooldown time from about 90 hrs to < 60 hrs.
Water flooding the reactor(s) must be done after an appropriate management of change process, and it must be done using the right procedures. For example, the reactor foundation must be checked to make sure it can tolerate the additional weight of the water. The correct processes must be put in place to separate water from the catalyst and to dispose of the water. The author highly recommends that personnel seek the advice of a licensor and catalyst handling company with extensive experience in water flooding reactors.
Application in plant units.
How long has it been since shutdown procedures were reviewed to improve the speed and efficiency of the shutdown and cooldown processes? Many refineries do not put much emphasis on this area.
What is the “size of the prize” for reducing shutdown time on hydrocracking and hydrotreating units. The cost to implement many of the ideas listed in this article is small, but the payout can be huge.
Are reactor cooldowns smooth and steady to help protect equipment?
Are the correct written procedures available to prevent nickel carbonyl formation and brittle fracture?
Has the hot strip time for reactor shutdowns been optimized? HP
LITERATURE CITED
Johns, J., “Preventing nickel carbonyl formation during hydroprocessing unit shutdowns,” Becht, August 9, 2023, online: https://becht.com/becht-blog/entry/preventing-nickel-carbonyl-formation-during-hydroprocessing-unit-shutdowns/
Jeff Johns has more than 35 yrs of experience in the petroleum refining industry. He was honored as a Chevron Hydroprocessing Fellow (Chevron’s highest technical recognition) for contributions to Chevron and to the industry. He has expert knowledge of hydrocracker and hydrotreater design/operation, optimization and troubleshooting, and has substantial experience in other key refinery processes. Johns has managed hydrocracking and hydrotreating technology in Chevron’s refineries worldwide where he has developed and implemented best practices and projects to improve safety, reliability and profitability. He was a member of the AFPM Q&A Panel in 2004 and directed multiple technology seminars as a member of the AFPM Q&A screening committee. Johns has also served on the Board of Directors for Advanced Refining Technologies (ART). He earned a BS degree in chemical engineering from the University of Utah and holds six patents in hydroprocessing technology.