The Media Is the Message

The advantages of using random packed ceramic heat recovery media with regenerative thermal oxidizers to improve efficiency and reduce operational costs

Over the last several years there have been widespread misconceptions about which type of heat recovery media (HRM) is best for heat exchange and horsepower usage when used with a regenerative thermal oxidizer (RTO), a type of air pollution control technology used to destroy volatile organic compounds (VOCs). The confusion occurs among end users in a variety of core industrial sectors including fertilizer, chemical, petrochemical, and ethanol manufacturing groups. The purpose of this article is to enlighten the end user that no one media can be considered a panacea; each application's HRM must be based upon its ability to interface with the specific industrial process, the type of RTO, and its mandated VOC-destruction efficiency. Hopefully after reading this article, the end user will understand that thermal efficiency and destruction efficiency are not inclusive of each other and that it is up to the RTO supplier -- not the media supplier -- to pick a media that will interface well with the hundreds of other components used to form a particular composite RTO, which ultimately has been purchased to destroy VOCs and not just to save energy.

Thermal Efficiency vs. Fuel Usage
While there are many technologies available to destroy VOCs, the one most commonly chosen is the RTO. It is selected for two main reasons. First of all, it is comparably inexpensive to operate -- close to 95 percent of the fuel used to elevate the fume to the VOC-combustion temperature is reused to preheat the incoming fume to within 5 percent of the combustion temperature. Simply put, it uses 95 percent less fuel than a common afterburner that recycles no heat. Secondly, it is a forgiving technology in that a wide variety of VOCs can be admitted without worry that if the process changes, there will be a need to change destruction technologies.

The technology, although simply described, is very dynamic in nature. By using a fan before or after the RTO to supply fume movement, the process gas is forced to enter the RTO's first recovery section by way of a valving system. The fume passes through the recovery chamber that contains a porous inert ceramic heat sink (the HRM), on its way to the combustion chamber. In the combustion chamber, its temperature is elevated to greater than 1,400 degrees Fahrenheit. Here the VOCs are oxidized and converted to water (H2O) and carbon dioxide (CO2). The clean gas exits the RTO through a second heat-exchange section also containing the HRM. At this time, close to 95 percent of the clean fume's 1,400-degree temperature is transferred to the HRM. The relatively cool, clean fume exits the RTO to the environment through a second valving system. After a predetermined period of time, the inlet and exit valving systems are switched, causing flow across the RTO to be reversed. The inlet chamber now becomes the exit chamber and the former exit chamber becomes the inlet chamber. Now the cold process fume is passed over the hot HRM, picking up close to 95 percent of the stored heat on its way to the combustion chamber. Now in the combustion chamber, the temperature only needs to be raised 5 percent by the burner to reach the preordained 1,400 degrees Fahrenheit. The HRM, by storing and then releasing combustion heat, reduces fuel-operating costs by approximately 95 percent.

While the common recycling efficiency touted is 95 percent, it must be understood that this percentage is a nominal rating that only occurs when there are enough combustible contaminants within the process fume stream to offset the fuel required to complete the final 5 percent elevation to combustion temperature, plus the fuel needed to overcome the heat loss of the RTO housing shell. When this occurs, the RTO mass balance is perfect (balanced mass flow) and the 95 percent nominal efficiency will be reached.

In most RTO fume process, the hydrocarbons (combustibles) within the fume are not adequate to self-support combustion. In this case, the burner must fire. Firing full bore, with little or no combustibles present and using a typical non-low-nitrogen oxides (NOX) burner, adjusted to 20 percent excess combustion air, close to 3 percent mass unbalance will occur. This will drop the nominal thermal efficiency to a true thermal efficiency of approximately 92 percent. While 3 percent loss does not sound like much as the RTO is still recovering 92 percent of the heat, in fact, the real difference between using 5 percent fuel and using 8 percent fuel is an addition of 60 percent more fuel. Using a low-NOX burner, which has a high excess air demand, will further increase fuel usage. The shock is not over -- when the fuel required to elevate the combustion air to combustion temperature and the loss of fuel heating value to water conversion are added in, the total fuel used will be twice that of a nominal 95 percent RTO. Over the last 30 years, there have been many lawsuits on this issue because the impact of the words "mass balance" on fuel usage were not understood.

Destruction Efficiency
As stated above, destruction efficiency and thermal efficiency are not inclusive of each other. In many cases, meeting thermal efficiency has a negative affect on VOCs conversion because each type of HRM has specific characteristics that must be designed around to meet destruction requirements. It is not a matter of one kind fits all.

For example, extruded monolith and/or structured media, used exclusively by some RTO manufacturers because of the overall cost savings in producing the device and lower particulate clog characteristics, comes in block form. Unlike saddles, which are random packed (dumped-in), they must be carefully placed within each recovery chamber. Both monolith and structured media are good heat exchangers, but only because of their thin web structure. They pick up heat and release it rapidly.

Unfortunately, to achieve the touted thermal efficiency, the length of time the process fume enters a particular chamber before chambers are transferred must be kept short -- between one and two minutes. This would be fine if the only goal was to design a heat exchanger; however, each time a chamber transfer takes place there is an unburned amount of VOC that is emitted into the atmosphere, which is commonly referred to as the "puff." In the case of the traditional two-chamber RTO, many marginally successful remedies have been employed to reduce the puff, but when greater than 98 percent destruction of VOC is required, cycle times must be stretched out to reduce the number of cycles, as each puff reduces overall destruction efficiency. Here is the paradox: stretching the cycle time to improve destruction efficiency lowers thermal efficiency, but the stretch also generates larger and larger HRM temperature swings as the elongated cycle time depletes the media of stored heat. In many instances, when using structured or extruded monolith media it is common to have cycle times beyond four minutes.

This generates an initial-to-final cycle-time temperature swing within the HRM of greater than 200 degrees Fahrenheit, which means the average bed temperature drops from a typical 725 degrees to 525 degrees Fahrenheit. Since most organic compounds begin to oxidize between 600 degrees and 800 degrees Fahrenheit, dropping the average bed temperature by 200 degrees Fahrenheit means that far less VOC destruction occurs in the bed, which must be made up for in the combustion chamber.

This is contrary to random-packed saddles HRM, which has cycle times that can be maintained for six minutes without excessive HRM temperature or efficiency loss simply because web thickness is four times greater than the monolith/structured media, as well as the additional volume that must be utilized to reach nominal 95 percent thermal efficiency. Saddles have a better heat-storage capacity. In fact, unlike monolith/ structured media, the temperature and turbulence created by saddles generate almost total VOC destruction before the fume reaches the combustion chamber. With the use of saddles, the combustion chamber size becomes less of an issue, as most of the VOC destruction is done within the saddle HRM bed. In contrast, when designing an RTO using monolith/structured media, it is of utmost importance to adequately size a combustion chamber because the excessively large entering fume temperature swings will wreak havoc on residence dwell times, combustion chamber temperature uniformity, and burner stability.

Since all the changing dynamics of an RTO are driven by cycle-time temperature swings, the greater the swings, the greater the challenge to generate stability. First the temperature within the RTO must be stabilized. Secondly, a consistent interface with the industrial process needs to be maintained since exhaust pressures and flows will vary widely if inadequately designed for.

Horsepower Requirements
There is a big difference in the horsepower (HP) utilized by an RTO using conventional 1-inch Interlox saddles and using monolith/structured media. Designing at 250-feet-per-minute (fpm) entering air velocity, with both RTOs achieving 95 percent nominal thermal efficiency, the one using monolith or structured media will generate one half the pressure drop as compared to the RTO using 1-inch saddles. Since pressure drop is directly proportional to horsepower, using the monolith media will require one half the HP compared to 1-inch saddles. One-inch saddles are used as an example only because they are the most commonly used saddle type with RTOs.

There is no question, because of its clear path area, pressure requirements, and requisite horsepower needed to push or pull flow, structured/monolith media is one-half that of conventional 1-inch ceramic saddles, given the same fixed-flow area. It is also true that 30 percent more height is required when using saddles. But in order to make use of the performance properties related to structured/monolith media, one must have a clear understanding of the constraints required of the RTO. In order to obtain the advantages associated with the use of structured/monolith media, the user will have to make some compromises.

Presently, there is a cost-effective option to structured/monolith media. The new generation random-packed ceramic saddles, with a bold scalloped slotted configuration, afford the end user the best of all worlds -- on an equal velocity basis, they have a pressure drop equal to that of structured/monolith media, and offer superior destruction efficiency. These saddles are specifically designed for RTOs. They are configured to generate the least pressure drop while providing the necessary time, temperature, and turbulence to destroy the majority of VOCs before entering the combustion chamber. Cycle-time durations of six minutes generate an HRM swing of less than 140 degrees Fahrenheit without a loss in thermal efficiency.

On an equal flow/area basis, the footprint of both RTOs will be the same. But since new-generation random-packed ceramic media, such as Cycle-Therm Inc.'s Cell-StoneĀ® ULTRA, costs one-fourth the price of structured media and can be quickly deposited in the chambers, it yields a much more attractive fabrication cost and retrofit expense with quicker installation time.

Since the pressure drop of any saddle decreases by the square of the face velocity, and the HP decreases by the cube of the face velocity, increasing the face area by 25 percent will reduce the HP requirements to move the process flow through both recovery chambers by 40 percent. In contrast, pressure drop through structured/monolith media is linear; therefore increasing the face velocity by 25 percent will only yield a 25 percent savings in HP.

Conclusion
Final and correct design results are, and always will be, subject to the RTO manufacturer's understanding of the relationship between the design parameters and the components. As discussed in this article, they include, but are not limited to, media configuration, mass flow imbalance, and burner configuration. Knowledge of destruction efficiency as it relates to time, temperature and turbulence, and valve cycle time also is important.

Other crucial factors to consider are the impact of the heat-up time, the size of the housing, the pressure drop, chamber transfer valve(s) leakage, unburned VOC recovery media void area, the weight of the media, and velocity through the media. Additionally, factor in the type of solvent, the shell insulation, the cycle time, the combustion chamber temperature, the design thermal efficiency, the exhaust temperature swing, the product limitations, and the means of testing.

The point is that an RTO is more than a "box of rocks." It is born from the interface of hundreds of components, interrelationships, and design constraints generated by certain specifications. If the end user does not have a clear understanding of these intricate interrelationships, the RTO will truly be reduced to a "box of rocks."

This article originally appeared in the 09/01/2004 issue of Environmental Protection.

comments powered by Disqus