NEWS RELEASE MAY 2016
Complex Unintended Consequences Obscure the Path Forward for Air Pollution Control
A small Ohio town no longer exists thanks to the unintended consequences of air pollution control. A nearby power plant spent hundreds of millions of dollars to reduce NOx. The catalyst not only reduced the NOx it converted SO2 to sulfuric acid. Within a few days, the acid deposition did such great damage to the buildings in the town that the utility agreed to buy the complete town and pay for relocations. In the ensuing decade, catalyst suppliers have redesigned their product to eliminate this problem.
New mercury regulations have such low emission limits that the instrument just to measure gaseous mercury can cost hundreds of thousands of dollars. Prior to issuing the regulation, EPA tested a number of stacks and found that all the mercury existed in gaseous form. Therefore, the regulations only required measurement of gaseous mercury. In response to the regulation requirements, power plants, cement plants and waste-to-energy plants embraced a two-step solution.
Step one was to convert the gaseous mercury to particulate mercury. Step two was to remove the particulate mercury. The end result is that if step one is very efficient and step two is not, there is lots of particulate mercury being emitted. Another unintended consequence is that particulate mercury will not travel far, whereas gaseous mercury can transverse the globe. Even though this problem has been evident for a few years, there is still no proposed change in the regulations.
The recent regulation of many pollutants combined with new technology which makes it possible to remove all the pollutants in one device has greatly increased the use of fabric filters. However, there has not been a recognition of what McIlvaine describes as “The importance of FIFO vs. LIFO in Dust Cake creation.”
Direct sorbent injection (DSI) and embedded catalyst dictate a new approach to bag cleaning. In addition to discrete particle capture, bag filters are being tasked with:
- Mercury removal
- Acid gas absorption
- Dioxin destruction or capture
- NOx reduction
The importance of the method of bag cleaning can be illustrated by use of the accounting approach to inventory. Two options are first in first out (FIFO) and last in first out (LIFO). If the price paid stays the same, the choice between the two accounting methods makes no difference. But, if the cost of recent inventory is greatly different than the past, then the accounting method makes a big impact on profits.
The capture of discrete particles is the equivalent of price parity. Let’s say that when you pulse a bag you are always discharging the latest particles to arrive and the remaining cake consists of the earliest. Since the ability of a matrix of dust particles to act as a filtration medium does not change, it does not matter which particles remain. In fact, maintaining a somewhat permanent layer of cake protects the fabric from wear. Also a more permanent cake provides higher dust capture. It has been shown that on-line cleaning results in some re-deposit of dust particles. But this is does not impact discrete particle capture efficiency.
The new paradigm with DSI is a big price difference. The newly arrived lime particle has the capability to absorb acid gases. The lime particle deposited earlier is already converted to calcium sulfate and provides no additional absorption capability. The semi-permanent cake layer is very undesirable for acid gas capture. Mercury re-emission is also a risk for an activated carbon cake which is semi-permanent. So it is very important to adopt FIFO and not LIFO.
This leads to the obvious question as to which are the best cleaning methods to achieve LIFO? The long running debate about surface filtration vs. depth filtration needs to be reviewed in light of FIFO. Also, the pulsing method itself needs to be reviewed. Do some methods result in more re-entrainment of particles in the previous cake than do others? Should more of the cake be removed with each pulsing?
It could be argued that the reaction takes place in the ductwork and not on the bag. But the big difference in performance of bag filters vs. precipitators with DSI proves that the cake absorption is substantial.
There may be lots of research on this subject but if so, McIlvaine would appreciate feedback on it. If there is not, it is an area deserving lots of attention.
Bag cleaning is also made more challenging by the increasing use of ceramic filter elements. The advantage of these elements is the ability to remove dust at 850°F. The older generation rigid ceramic has been replaced by ceramic fiber media which can be pulsed. However, this media cannot necessarily be pulsed with the identical system used for synthetic bags. An alumina refinery in Australia was having cleaning problems with a ceramic filter. Pentair Goyen analyzed the situation and provided a more robust pulsing system. This solved the problem.
Ceramic, glass and even synthetic media are incorporating catalyst in the media to reduce NOx or oxidize dioxins. Do these designs require a different cleaning approach? The catalyst in the Clear Edge design is not on the surface. So, the dust cake will not affect performance except if it causes maldistribution of the gas. If more gas flows through one area than another, the reactivity of the system is reduced.
A broader subject is the whole approach to cleaning. High pressure/low volume is the most popular option. Does capture of these other pollutants open the door for high volume /medium pressure or even for reverse air cleaning?
The potential for the one-stop shopping is great. Costs of pollution control can be reduced for new installations. The small footprint makes a big difference in the cost of upgrading existing plants to meet new air pollution rules. It is, therefore, important to understand and then maximize FIFO potential. McIlvaine will be interviewing experts in the various niches to shed more light on this. The results will be published in:
44I Power Plant Air Quality Decisions (Power Plant Decisions Orchard)