BRENNEN

Industries / Food, farming and forestry / Forestry and logging / Timber and sawmills / Lumber treating

Testing and standards

Setting how deep preservative must go and how much must stay in the wood for each use, checking treated wood against those rules, and learning from decades-long field tests how long it lasts.

6

Milestones achieved, from 1904 to 2016.

1

To watch for: announced, not yet confirmed outside.

2

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1904

    A shared rulebook begins

    The Wood Preservers' Association first met in 1904, when US railroads were using about 110 million crossties a year, and took the name American Wood Preservers' Association in 1912, by the account of Roy Baechler and Lee Gjovik of the US Forest Products Laboratory (FPL). Now the American Wood Protection Association (AWPA), it calls itself "the principal Standards-writing body" for the preservatives and methods that protect wood, and the major model building codes cite its standards. A new preservative enters them in two steps: AWPA committees first list it as a potential preservative after reviewing laboratory and field tests, then set the treatment level for each product (FPL's Wood Handbook).

    foresthistory.orgawpa.comresearch.fs.usda.gov

    Credit

    AWPA

  2. 1938

    Treated stakes in the ground

    In 1938 FPL and its cooperators began treating stakes of southern pine sapwood, 2 by 4 by 18 inches (about 5 by 10 by 46 cm), with different preservatives and setting them in the ground at the Harrison Experimental Forest at Saucier, Mississippi, with matching sets in Wisconsin, Louisiana, Florida and the Panama Canal Zone. Inspectors pull each stake and grade decay and termite attack on a scale from 10 (sound) to 0 (failed). In FPL's 2011 progress report, untreated stakes lasted about 1 year on average in the Canal Zone, 1.8 to 3.6 years in Mississippi, Florida and Louisiana, and about 6 years in Wisconsin; after 24 to 61 years in Mississippi, 20% of stakes treated with chromated copper arsenate (CCA) at 0.44 lb/ft³ (about 7 kg/m³) or more had failed.

    LessonField data take decades and cannot be redone; FPL notes that many old plots "cannot be duplicated or replaced" because their preservatives have since been discontinued or banned.

    research.fs.usda.gov

    Credit

    Forest Products Laboratory; report by Bessie Woodward, Cherilyn Hatfield and Stan Lebow

  3. 1960s

    A Nordic treated-wood mark

    Denmark, Finland, Norway and Sweden started the NTR system in the 1960s. The Nordic Wood Preservation Council classifies treated wood, approves preservatives and has treating plants checked by a third party; the scheme has since spread to the Baltic states and Poland. By the European Wood Protection Association's count, 108 companies in 9 countries hold certificates to use the NTR mark, and the most used classes are NTR A for ground contact and NTR AB for above ground.

    ewpa.org

    Credit

    Nordic Wood Preservation Council

  4. 1999

    Wood sorted by exposure

    AWPA adopted its Use Category System, standard U1, in 1999 and has revised it every year since. It sorts every use of treated wood by how harsh the exposure is, from UC1 (indoors and dry) through UC3B (outdoors above ground) and three grades of ground contact (UC4A to UC4C) to UC5 (sea water, by region), with separate categories for fire-retardant wood indoors and outdoors (UCFA and UCFB). The use category on a board's end tag sets how much of each preservative it must hold.

    awpa.comawpa.com

    Credit

    AWPA Technical Committee T-1

  5. 2003

    Untreated framing reversed

    Boron treatment of New Zealand framing began in 1952, and until 1992 most house framing there was boron-treated radiata pine. From 11 September 1995 the standard NZS 3602 allowed untreated framing if it was kiln-dried and stayed at or below 18% moisture in service, and the building code's compliance document B2/AS1 cited it from 28 February 1998 (Standards New Zealand). The change rested on "the belief homes would stay dry", in the words of New Zealand's Bioeconomy Science Institute. From 1998 to April 2004 homes were often framed with untreated kiln-dried pine, and where their walls leaked, decay was "more likely" to be "both severe and widespread", by the government's guide to repairing leaky buildings. A Building Industry Authority (BIA) directive of 9 March 2003 required treated timber for all consents issued from 1 April 2004, by Standards New Zealand's account; the BIA's decision announced on 23 December 2003 raised exterior wall and subfloor framing from H1, which resists insects only, to H1.2, which also resists decay, and still allowed untreated kiln-dried pine and Douglas fir in lower-risk parts. The BIA put the cost at about NZ$500 for an average house. Since 2011 only boron treatments can be used for H1.2 framing (BRANZ).

    LessonA rule that counts on a wall staying dry needs a margin for the wall that leaks.

    standards.govt.nzbuilding.govt.nzbioeconomyscience.co.nzscoop.co.nzbranz.co.nz

    Credit

    Building Industry Authority; Standards New Zealand; BRANZ

  6. 2016

    Deck joists moved up

    Before 2004 much US deck framing was CCA-treated to the ground-contact level, UC4A. When makers stopped producing CCA for homes at the end of 2003, framing was often treated with the replacement preservatives only to the lower above-ground level, UC3B. By the end of the next decade builders reported joists decaying where the deck boards sit on them; one builder quoted by Frank Woeste said framing was "lasting 7 to 10 years before significant decay has shown up", against two to three decades expected of CCA decks. In 2016 AWPA added a footnote to U1: joists and beams "shall be treated to requirements for UC4A when they are difficult to maintain, repair or replace and are critical to the performance and safety" of the structure.

    LessonA lower treatment level can pass short tests and still show its failures a decade later in service.

    jlconline.comawpa.com

    Credit

    AWPA; account by Frank Woeste in the Journal of Light Construction, October 2024

Watch for

  1. Pole-top decay scored by AI

    ComEd, the Chicago-area power utility, says a computer vision model it completed in early 2025 scores the decay visible in drone photos of the tops of its wooden distribution poles and ranks which of its more than 1.3 million poles need replacing or reinforcing. It says the model "has already reached 86 percent accuracy compared to human expert agreement", and that six retired poles the model flagged went to its forensics lab to be cut open; it published no results from them.

    Achieved whensomeone outside the company compares the model's scores with sectioned poles or hands-on inspections across a published sample and reports misses and false alarms.

    poweringlives.comed.com

    Claimed by

    ComEd, April 2026

Open

  1. 01

    Stake test records as data

    A machine-readable dataset of every stake group in FPL's stake-test progress reports, with preservative, retention, site, year installed and each inspection's ratings, plus survival curves (the share of stakes still in service by year) for each preservative and retention.

    The brief
    Credit · unclaimed
  2. 02

    Store-bought lumber, assayed

    A public audit of treated lumber as it is sold: buy boards at home centers and lumber yards in several regions, cut cores the way AWPA's standards prescribe, have an accredited laboratory measure the preservative in each, check penetration with AWPA's color indicator, and set each result beside the retention and use category printed on the end tag. Retention is the amount of preservative held in a volume of wood, in kilograms per cubic meter (kg/m³) or pounds per cubic foot (lb/ft³).

    The brief
    Credit · unclaimed

Open 01 · the brief

Stake test records as data

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A machine-readable dataset of every stake group in FPL's stake-test progress reports, with preservative, retention, site, year installed and each inspection's ratings, plus survival curves (the share of stakes still in service by year) for each preservative and retention.

Why it is still open

FPL's 2011 report gives its results only as 76 tables in a 120-page report, and says that for plots that were closed "inspections were discontinued, and the database maintained in archival records". The IRG/WP Durability Database, an open platform run by the International Research Group on Wood Protection for raw field-test results, holds 178 records, all PDF files, the newest numbered from 2016; 10 of them are for preservative-treated wood (count of its index, November 2026). Searches of Zenodo and the Forest Service Research Data Archive found no stake-test dataset.

research.fs.usda.govirg-wp.com

Built with what exists

pdftotext and Camelot to pull tables from the PDFs; an AI document reader for the older scanned progress reports, with every number checked against the page (the Note "Seven more flagged" covers an AI reviewer of long documents); Python with pandas and the lifelines survival-analysis library; Zenodo to host the data, and the IRG/WP database's own field-test template for submission.

  1. First step

    Extract Table 77, FPL's summary of the Mississippi results for preservatives in general use, and reproduce its average life column from the detailed tables it cites.

  2. Done looks like

    Every table in the 2011 report published as CSV with a column-by-column data dictionary; the dataset reproduces the report's average life for at least 20 preservative and retention groups; survival curves for CCA, creosote and pentachlorophenol at Saucier published alongside.

Open 02 · the brief

Store-bought lumber, assayed

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public audit of treated lumber as it is sold: buy boards at home centers and lumber yards in several regions, cut cores the way AWPA's standards prescribe, have an accredited laboratory measure the preservative in each, check penetration with AWPA's color indicator, and set each result beside the retention and use category printed on the end tag. Retention is the amount of preservative held in a volume of wood, in kilograms per cubic meter (kg/m³) or pounds per cubic foot (lb/ft³).

Why it is still open

Treated wood is checked at the plant, by the treater and by third-party inspection agencies that sample whole charges; a buyer sees only the end tag, which AWPA tells homeowners to read for "AWPA U1", the use category and the retention. The one store-bought sample found is a 2012 study by Patricia Lebow of FPL with Adam Taylor and Timothy Young of the University of Tennessee: 60 southern pine 2x4s from two Knoxville retailers, all tagged 0.06 lb/ft³ (0.96 kg/m³), with mean copper retentions of 0.052 and 0.058 lb/ft³ and board-to-board variation of 27% and 50%. They used it to build a statistical tool for treaters, not to survey the market. Searches in November 2026 found no wider independent survey.

research.fs.usda.govawpa.com

Built with what exists

A plug cutter or increment borer; the chrome azurol S spray that turns copper-treated wood deep blue (AWPA A3); X-ray fluorescence (XRF) analysis of milled cores at an accredited laboratory; the lower prediction bound that Lebow, Taylor and Young published, worked in Python; a phone camera for every end tag; gloves and a dust mask, with cutting done outdoors.

  1. First step

    Buy 20 boards with the same tag from each of two stores, cut the samples, and get quotes from two accredited laboratories for copper analysis.

  2. Done looks like

    At least 300 boards from at least 30 stores in at least 5 US states or 3 countries, each with its tag photo, measured retention and penetration, published openly with the share that falls short of the retention on its tag.