Keep out: certain chemical preservatives are unwelcome contaminants in the mulch and wood fuel markets.

Markets for recycled construction and demolition (C&D) wood include mulch or wood fuel. Both options minimize landfill disposal and serve as substitutes for new wood materials.

However, the presence of preservatives within C&D scrap wood greatly degrades the quality and thereby limits the potential of this valuable material.

PROBLEM PRESERVATIVES. Preservatives are added to wood intended for outdoor applications (e.g. fences, decks, docks, etc.) to provide protection against fungi and termites. Treated wood can also be used indoors in areas where the wood is in contact with the foundation of a building or in high termite hazard areas. The amount of chemical added to wood varies depending upon its intended use. The amount added, or "retention level," is described in pounds of chemical added per cubic foot of wood (PCF).

Preservative-treated wood can be broadly separated into two general categories--oil-borne preservatives and waterborne preservatives. Oil-borne preservatives use an oil or organic solvent as the carrier during the pressure treatment process, whereas waterborne preservatives are dissolved in water that then serves as the carrier.

Oil-borne preservatives, such as creosote and pentachlorophenol, are traditionally used for heavy-duty industrial applications, such as utility poles and railroad ties. Because of their large dimensions, wood treated with oil-borne preservatives can be readily identified and removed from the recycling stream.

Waterborne preservatives, on the other hand, are used to treat not only industrial products but also to treat products used in residential areas, such as lumber, timbers and plywood. As a result, this portion of treated wood can be easily commingled during disposal with wood that is untreated. One primary challenge when recycling C&D wood is thus identifying and sorting out preservative-treated lumber, timbers and plywood.

Currently, several different types of preservatives are used for treating lumber, timber and plywood. The most common of these preservatives include CCA, which contains chromium, copper and arsenic; ACQ, which contains copper; CBA, which contains copper and boron; ACZA, which contains copper, zinc and arsenic; and borate-treated wood, which contains boron. Among the elements within the common wood preservatives, arsenic is characterized by the strictest regulatory standards, chromium and copper are the next strictest, followed by boron and zinc.

For example, soil clean-up target levels (SCTL, which are typically used as a guideline to determine if a recycled material can be land applied, such as in the case of mulch) are generally lowest for arsenic by roughly a factor of 100 relative to that for copper.

If the goal is to meet Florida SCTL residential guidelines for land application of recycled materials, the amount of arsenic-treated wood (either as CCA or ACZA) that can be commingled with untreated wood must be less than 0.05 percent, which is less than 1 pound of arsenic-treated wood per ton of wood processed.

For copper-treated wood (either as ACQ or CBA) the guidelines are less strict, permitting for up to 2 percent to 3 percent non-arsenical copper-treated wood, or between 45 to 60 pounds of copper-treated wood per ton of wood processed. Because of the high SCTLs for boron, there are no limits on the amount of boron-treated wood.

Because of the strict regulatory levels for arsenic, compliance with the regulatory guidelines is difficult. We estimate that at the Florida recycling facilities we have observed since 1996, the amount of arsenic-treated wood in incoming loads ranges from 6 percent to 30 percent.

Identifying arsenic-treated wood is not always easy. This dilemma has led wood recyclers to ask, "How can C&D wood be recycled while minimizing contamination from preservative-treated wood?"

This question is being addressed through a governmental, university and industry collaborative effort funded through the Florida Department of Environmental Protection (FDEP) Innovative Recycling Grants Program. The fund recipient was the town of Medley in Dade County, Fla. Medley has facilitated the cooperation of University of Miami and University of Florida researchers to work with Florida Wood Recycling to document the efficiency of different sorting methods to identify treated wood within C&D scrap wood.

THE SORTING SOLUTION. This study has documented that a multi-tiered approach is best for identifying preservative-treated wood debris. The first line of defense is to inspect loads as they enter the facility.

Florida Wood Recycling, a mid-sized recycling facility located in Medley, accepts yard debris and source separated C&D wood (C&D wood that has not been commingled to a great extent with other components of C&D debris). As part of this stud> the facility has initiated the processing of commingled C&D debris containing wood (a.k.a. commingled C&D wood).

For the source-separated C&D wood, Florida Wood Recycling inspects each load as it enters the facility. Inspection has been based upon visual identification and knowledge concerning the origin of the wood. Visual identification relied upon many different factors, including looking at the load to evaluate if it contains remnants of an exterior structure, such as portions of a fence or a dock, indicating it likely contains treated wood.

Other criteria include observing the general dimensions and color of the wood and looking for end tags. Wood characterized by very large dimensions had likely been used for industrial applications and are almost exclusively treated and should not be recycled as mulch. Examples include railroad ties that are 8 inches by 8 inches by 3 feet or more and utility poles (typically 1 foot in diameter or more).

In some cases, landscape timbers that are also typically treated can be identified by their shape, which in many cases is characterized by rounded edges for decorative purposes.

End tags provide another important clue as to whether or not wood within a load is treated. End tags typically list the type of chemical contained in the wood.

In some cases a load containing treated wood can be identified by observing the color of the wood, which, if not treated, typically has a light yellow hue. If treated with a copper containing preservative (e.g. ACQ, CBA, CCA, and ACZA) the wood would be characterized by an olive color that is faint for lower retention level wood and very distinct for wood treated at high retention levels.

If the wood is incised, it is also treated. Incising is a process by which uniform cuts are made in the wood to improve the penetration of the preservative during treatment. Incising is typically used for denser wood species, such as Douglas Fir, which is primarily used in the Western United States.

Close observation of incoming loads of source-separated wood has resulted in notably "clean" wood at Florida Wood Recycling. Of 10 tons of wood sorted, only 100 pounds (0.5 percent) was treated. This fraction of treated wood is considered to be exceptionally good quality at the first line of defense. After the initial inspection at Florida Wood Recycling, the wood is further evaluated once it is unloaded, after it has been moved to the mulch supply pile and once it is on its way to be chipped.

The study has also shown it is much more difficult to control the quality of the incoming scrap wood stream if the wood is commingled with other C&D material, such as concrete, drywall, tile and the like. Commingled C&D loads at Florida Wood Recycling are first inspected upon arrival and the wood is then sorted out using a picking line.

Of nine tons of wood inspected from the picking line, almost 1,400 pounds (or 8 percent) was treated. Of the 1,400 pounds, about 10 percent was treated with creosote, 67 percent was treated with CCA, and 23 percent was treated with copper-based alternatives. The net fraction treated with CCA was 5 percent. Although some plywood was observed to have been CCA-treated, the vast majority of the CCA-treated wood was in the form of lumber and timbers.

TOOLS OF THE TRADE. Because of the difficulty of visually inspecting commingled C&D materials for the presence of CCA-treated wood, removal of arsenic-treated wood at C&D recycling facilities accepting commingled C&D debris will thus depend more heavily upon "back-end" inspection--inspection after the wood is sorted out on a picking line.

However, visually identifying preservative treatment for wood sorted on a picking line is much more difficult than for source-separated C&D scrap wood. The difficulties are two-fold.

First, in a picking line sort, the wood is broken to smaller pieces. Overall, data collected through this study shows there were about 360 pieces of wood per ton of source-separated C&D wood. This number doubled to roughly 730 pieces of wood per ton after sorting on the picking line.

Furthermore, when wood is commingled with other components of the C&D material stream, it is covered with dirt and dust from crushed concrete and drywall. As a result, visual identification of treated wood becomes extremely difficult. One can no longer identify if the wood was part of a fence or other outdoor structure because the wood has been broken to smaller pieces and the green hue of the wood, in particular for wood treated at lower retention levels, is no longer visible.

In the case of wood recovered from commingled C&D debris, it is extremely important to sort "after the fact" using augmentation methods. Augmented sorting is defined as traditional visual sorting (as described above) plus an additional technology.

The technologies evaluated at Florida Wood Recycling include the use of PAN (pyridylazo and naphthol) stain that, when applied to wood, causes a distinct color change if the wood is treated with copper. Or, the use of hand-held X-ray fluorescence units can identify as much as 15 different metals in wood, including chromium, copper, arsenic and zinc (Blassino et al. 2002; Solo-Gabriele et al. 2004).

The PAN stain was found to work well for cases where the wood was free from excessive dust and dirt and for wood that was not overly wet. This limitation may present a problem for facilities that are not paved (which increases dirt) and if C&D debris is sprayed with water for dust control.

Labor for PAN-augmented sorting was roughly two-to-three times that for visual sorting. The benefits of using PAN stain are most apparent for loads containing a significant fraction of treated wood. When considerable amounts of C&D wood are treated within a load, visual sorting is not as efficient resulting in up to as much as 50 percent of the wood being incorrectly categorized as untreated.

At a minimum, PAN stain should be used to sort loads suspected of containing on the order of several percent of treated wood. Sorting with the hand-held XRF units (in this case provided by Innov-X Inc., Woburn, Mass.) have a significant advantage over the PAN stain in that XRF units can be used to specifically identify arsenic vs. copper.

Furthermore XRF traits were not subject to interferences from overly dirty or damp wood. It takes about four to five seconds to inspect a piece of wood using XRF units, increasing sorting times by up to 10 fold; however in this case the efficiency of sorting reaches nearly 100 percent.

The research has found that XRF units are excellent tools for training and for spot-checking wood, in particular for loads that may look suspicious. For example, at Florida Wood Recycling a load of wood characterized by very large dimensions was brought to the facility. Such a load would have been rejected based upon visual considerations alone. However, the load was spot checked with the XRF units and was found to be untreated, so the load was accepted.

The utility of the XRF units is believed to lie in the inspection of C&D loads that are difficult to identify visually. The hand-held analyzers have been found to be very effective for identifying treated wood after a picking-line sort. The XRF units also have the potential to be incorporated into an automated inspection system using conveyors in a configuration such that each piece of wood is inspected without excessive added labor costs. Such automated systems may be cost effective for large facilities that sort commingled C&D materials.

FORWARD THINKING. Preservative-treated wood represents a challenge to C&D wood recyclers in that such wood must be carefully sorted out. "Front-end" management of the material stream at the incoming side of the facility is particularly effective for source separated C&D wood.

For commingled C&D wood, quality control will rely more heavily on "back-end" sorting using augmentation methods (e.g. a stain or XRF technology) to properly identify treated wood. One sense of relief comes from the fact that as of 2004, arsenic-treated wood was banned from most residential uses (U.S. EPA 2002).

Already, changes have been observed in the replacement of arsenic-treated wood with non-arsenic copper-treated wood within construction loads. In demolition loads, however, because of the long service life of treated-wood products, the industry will continue to observe gradual increases in the quantities of arsenic-treated wood for the next 10 to 15 years. After this point, quantities are anticipated to decline slowly.

For the near future, arsenic-treated wood will continue to represent a challenge to C&D recyclers nationwide. These challenges can be overcome to a great extent through a combination of visual sorting and additional tools to identify treated wood within the C&D waste stream.

Authors Helena Solo-Gabriele and Gary Jacobi are professors at the University of Miami, Coral Gables, Fla. Brajesh Dubey and Timothy Townsend are professors at the University of Florida, Gainesville, Fla.