Removing Contaminants From Spent Catalysts

September 1, 2013

catalyst recovery - part 3


Robert T. Jacobsen, Sabin Metal Corp.
Part 3

THE FIRST ARTICLE IN THIS SERIES ( 1) outlined the basics of recovering and refining the remaining precious metals from spent chemical, petrochemical and hydrocarbon catalysts and how to evaluate, select and work effectively with a precious-metals refiner. The second article ( 2) discussed in more detail the sampling procedures employed by refiners to evaluate the remaining precious-metals content of large lots of spent catalyst materials. This is a critical issue, because ultimately, the value of the precious metals returned to the owner is predicated upon meticulous and accurate sampling.

A key consideration in maximizing precious-metals recovery is the removal of contaminants from the entire lot prior to actual sampling. While this was covered briefly in the previous article, it warrants further discussion here because of its importance to — and influence on — the overall recovery and refining process.

Precious-metal-bearing catalysts used to facilitate reactions in hydrocarbon processes are subject to harsh operating conditions during their lifetimes and must be periodically regenerated, either while in operation in a continuous moving-bed catalytic process or removed and shipped off-site, to eliminate contaminants accumulated during processing. Eventually, however, they lose their effectiveness to the point where they must be replaced by fresh catalyst materials.

Precious-metal-bearing catalysts may exhibit high losson-ignition (LOI) characteristics due to moisture, volatile hydrocarbons and contaminants such as sulfur and carbon. Removal of the substances is critical to the downstream sampling process and to the efficiency, safety and environmental compliance of the subsequent refining processes.

Any sampling system, especially one designed to handle large lots, requires materials that are free-flowing. This is often not the state of the catalyst when it is removed from the process unit. Instead, the spent catalyst is often moist or sticky and will not flow through an automatic sampler. Heavy carbon deposits, known as coke balls, are often present, and can be as large as a grapefruit. Pre-burning will remove these physical impediments to achieving a representative sample.

The analysis of the sample’s precious-metal content must be adjusted for any changes in the material’s weight during handling. This is usually accomplished by determining the loss of weight upon ignition of a sample at a high temperature. Samples should be drawn as close to the net weighing of the lot as possible and in an undisturbed state.

This is in contrast to the sample drawn for preciousmetal analysis, which is subjected to many conditions that the bulk of the lot (sometimes called the reject) is not. For example, the sample material will be repeatedly ground and exposed to significant airflow during screening; these and other handling conditions during sampling can cause weight loss through evaporation of volatiles or weight gain due to the absorption of moisture from the atmosphere.

An LOI sample is either ignited immediately ( at approximately 900 C) or placed in hermetically sealed containers for later analysis by both the customer and the refiner. The LOI of the sample is expressed in weight percent. This factor is then applied to the net weight of the lot to calculate the ignited, or “settlement,” weight of the lot. Later, an analytical laboratory will subject a quality-control sample to the same conditions and determine the precious-metal content on an ignited basis.

Even if the representative sample is taken, the presence of volatile components and/or moisture may cause weight change during transit to the laboratory. The loss on ignition procedure described above will compensate for this possibility.

LOI determination is inherently less precise and less accurate than a settlement assay, and its variance must be as low as possible when samples are sent to the assay laboratory. The average precision of an LOI measurement is about 2% relative, although the range can be substantially more (up to 10%).

The table illustrates the effect of pre-burning on the precision of the entire settlement, even if the LOI can be determined to +2%. When the entire lot of catalyst is pre-burned, the LOI of the remaining material is reduced to about 1.5% of the lot. The uncertainty is thus 2% of 1.5% (0.03%), so the settlement weight is accurate to +18lb (60,000 X 0.0003).

If, however, the LOI is determined on the catalyst as received, and it is still possible to achieve 2% relative precision (which is doubtful when the LOI is as high as 40%), the settlement would only be accurate to +800lb (0.008 X 100,000). Of course, this effect would be much larger if the accuracy of the LOI determination were lower. These uncertainties are separate from, and in addition to, the sampling precision and the analytical precision.

catalyst recovery - part 3

Figure 1. A rotary kiln at the refiner’s facility removes contaminants from spent catalyst materials prior to sampling.


Pre-burning is often done in an indirectly fired rotary kiln (Figure 1). The typical rotary kiln will remove carbon and sulfur contaminants at catalyst throughput rates of 300-1,000lb/h. The contaminants may also be removed in a multiple-hearth furnace or fluidized-bed furnace. An indirectly fired rotary kiln is more versatile and controllable than other methods, in that it is easier to control the throughput time (the time in a reaction chamber), the temperature and the oxygen supply to the reaction.

Also important– at least from a financial perspective– is where the contaminants are removed. Overall refining costs can be reduced significantly when the kiln is located at the refiner’s facility. Many catalyst users must first ship their large lots of spent catalysts (as much as35,000-500,000lb) to an independent facility. There, strip burning removes hydrocarbons and coke burning removes carbon. In addition, another furnace may be required to dry fine particulates and other materials to eliminate moisture content.

Added turnaround time and additional costs are two major disadvantages of off-site strip and coke burning of spent catalysts. If these capabilities are not available at the refiner’s facility, catalyst users must pay substantial transportation charges to ship the catalyst to an independent off-site facility, where it might remain up to a month before it is processed. Then it would have to be shipped to the refiner to start the actual sampling, analyzing, recovery, and refining process.

During this time, the precious metals are unavailable to the catalyst user, and new metal must be acquired at current market prices and lease rates. At the current platinum value of approximately $900/oz and today’s platinum lease rate of 7% for a typical hydrocarbon catalyst containing 0.3% platinum, the finance charge is $0.052/lb. For a typical shipment of 100,000-200,000lb, this represents a cost of $5,000-10,000/wk of delay.

In addition, the indirectly fired kiln used for contamination removal must incorporate extremely accurate, multiple temperature zones programmed for the specific catalyst type and contaminants present. And, the kiln should be part of a system that incorporates downstream air-pollution-control equipment such as scrubbers, baghouses and incinerators to comply with applicable environmental regulations.

Perhaps the most important advantage of having the refiner handle the pre-burning in-house is the control the refiner has over every catalyst lot. This accountability eliminates the possibility that your material could be mixed with unrelated materials from another organization. When that happens, it is possible to accurately calculate its actual value.

  1. Jacobsen, R.T., “Catalyst Recovery–Part 1: Recovering Precious Metals From Spent Catalysts–The Basics. “Chem. Eng. Progress, 101(2), pp. 20-23 (feb. 2005)
  2. Jacobsen, R.T., “Catalyst Recovery–Part 2: Ensure Fair Market Value Through Accurate Sampling,” Chem. Eng. Progress, 101(3), pp. 22-27 (Mar. 2005).