Life cycle assessment

What if the next building did not begin with a new tree?

At Pengembak Beach, Alexis Dornier designed a pavilion around what was already there: the trees, the rocks, the shade, and the shape of the site.

Kaltimber built it from reclaimed ironwood, commonly called ulin, recovered from old, torn down buildings and structures in Kalimantan.

To understand the environmental impact, we conducted a Life Cycle Assessment (LCA) comparing the pavilion, as built using ironwood, with an identical, modelled pavilion, built using newly harvested, common tropical hardwood such as bengkirai. Over a 50-year life cycle period, the newly harvested hardwood pavilion would have produced a climate impact almost five times larger.

Reclaimed
ironwood

12,459 kgCO₂e, 50 years

Newly harvested
bengkirai

62,205 kgCO₂e, 50 years

The
difference

5× the climate impact

The newly harvested scenario produced 49,746 kgCO₂e more than the reclaimed one.

The Second Growth pavilion at Pengembak Beach: reclaimed ironwood posts, a slatted roof and a timber deck built around the existing trees, with the sea behind. — **“Second Growth”** by Alexis Dornier for Kaltimber, built from reclaimed ironwood. Structure by Gonçalo Miranda.
Jia Curated 2026 · Booth 19 · August 13–17, 2026

The question behind the pavilion

Timber is often thought of as being a responsible natural material.

Compared with cement or steel, it seems to offer an obvious environmental advantage: it grows, stores carbon, and can eventually be replaced by another tree.

But “timber” is not one environmental category. Different types of timber have very different environmental and ecological impacts. Where it comes from matters.

Reclaimed ironwood, ulin, has already been harvested, in some cases over one hundred years ago. The material exists, and using it again does not require another tree to be cut down.

Newly harvested timber is different, even when it comes from a legal concession, even when harvesting is selective, and even when the forest is expected to regrow. A new tree still has to be removed, the carbon stored underground, in soils, and above ground, in biomass still carry a cost in the model, and the material needs to be replaced when its service life ends, which is much more frequent than we often think.

That is what we explored through the work for this pavilion.

We compared the pavilion as built, using reclaimed ironwood, ulin, against an identical, modelled pavilion constructed in newly harvested tropical hardwood, bengkirai.

Same size. Same design. Same structure. Same purpose.

The difference was how the timber was sourced.

Visitors under the Kaltimber pavilion, a reclaimed ulin roof and deck set among the beach trees, with a carved ironwood post in the foreground. — The pavilion at Pengembak Beach during Jia Curated 2026.

New to life cycle assessment? Start here

A life cycle assessment, or LCA, counts the environmental impact of a building or a product across its whole life. It starts where the material comes from and ends when the building is taken apart.

The stages are standardised, which is what makes two options comparable. Both scenarios on this page are measured against the same set.

Diagram of the life cycle assessment stages, from A1 to A3 product, A4 to A5 construction, B1 to B7 use, and C1 to C4 end of life. — The life cycle stages used in this assessment.

Each stage carries a code:

A1–A3 making the material, from extraction to the factory gate
A4–A5 getting it to site and putting it up
B1–B7 the years in use, including energy, maintenance and replacement
C1–C4 taking it down, transporting the waste, and disposing of it

An LCA is only as reliable as the assumptions behind it and ours are available further down this page.

The result

Scenario	Total impact over 50 yearskgCO₂e	Impact per m²kgCO₂e
Reclaimed ironwood	12,459	161.8
Newly harvested bengkirai	62,205	807.9

The newly harvested scenario produced 49,746 kgCO₂e more than the reclaimed one.

That is almost five times the climate impact.

The comparison includes material supply, transportation, construction, replacement, operational energy, and end of life.

Stage by stage

Life cycle stage	Reclaimed ironwoodkgCO₂e	Newly harvested bengkiraikgCO₂e
A1–A3 Materials and production	−10,633	7,118
A4–A5 Transport and construction	1,682	1,731
B1–B7 Use, maintenance, replacement	15,936	45,530
of which replacement (B4–B5)	0	29,595
of which operational energy (B6)	15,936	15,936
C1–C4 End of life	5,475	7,826
Total, 50 years	12,459	62,205
Per m² over the lifetime	161.8	807.9

All figures kgCO₂e. The production stage for reclaimed ironwood comes out below zero because the assessment includes the biogenic carbon already stored in timber that exists. Operational energy is identical in both columns, so it cancels out of the comparison.

Method, assumptions, and boundaries

This life cycle assessment follows EN 15804 using the Level(s) life-cycle carbon method over a 50-year study period.

The pavilion is modelled as two timber routes:

approximately 6.0 m³ of reclaimed ironwood
the entire pavilion modelled in newly harvested bengkirai for the alternative scenario

The reclaimed scenario assumes recovery from existing structures, a 50-year service life, no replacement during the study period, reuse at end of life where applicable, and incineration for the remainder.

The newly harvested scenario assumes legal sourcing, a 20-year service life, and replacement over the 50-year study period. Illegal supply is not included.

Both scenarios assume air-dried timber rather than kiln-dried timber. Biogenic carbon is included and reported separately.

The system boundary includes:

material production
transportation
construction
operational energy
replacement
deconstruction
waste transport and processing
end-of-life disposal

Land-use change is modelled as forest converted to cropland and palm oil, using a difference of 102 Mg C per hectare in soil organic carbon and 140 Mg C per hectare in standing biomass.

This is a defined modelling pathway and a significant driver of the result. Full assumptions, data sources, calculations, and interpretation are available on request.

Where the difference comes from

The first assumption might be that the difference comes from cutting, milling, and transporting the timber.

Those activities matter. They are not the main reason the results diverge.

The largest difference comes from the carbon emitted in the land use change associated with producing new timber, and from having to replace that timber in 20 years.

Impact category	Reclaimed ironwoodkgCO₂e	Newly harvested bengkiraikgCO₂e
Fossil emissions	18,692	22,873
Biogenic carbon	−6,234	0
Land-use change	1	39,333

The newly harvested scenario assumes legal sourcing and does not include illegal supply. But here is what happens with the forest.

Land use change

Forest in Indonesian Borneo. — Indonesian Borneo.

The logging license

The government issues a selective logging license to take only the large, high-value trees. From the air the forest still looks intact. By the standards of Indonesian forestry law, this is the responsible version.

The road

Extracting a few large trees requires haul roads and landings that damage far more than the trees taken. Those roads are permanent and unpoliced. The trees leave. The roads stay.

The degradation

Overcutting, fire and encroachment accumulate, often in concessions the holder has abandoned or stopped monitoring. Replanting obligations get neglected, which helps make the forest “unproductive” and more likely to be reclassified.

The reclassification

A 2021 regulation lets palm oil permits be issued on forest area that’s no longer productive, explicitly including abandoned logging concessions. A release decree converts the land out of the state forest zone, after which plantation permits are entirely lawful.

Gone

A timber utilisation permit then allows the company to fell and sell everything remaining, often funding the plantation itself. Roughly a third of Indonesia’s cleared old-growth is never planted with anything, which suggests the timber was the point all along.

The first cut is the one that opens the forest. Everything after it follows a road that was built legally, to take a tree that was bought legally.

Within the selected land-use-change pathway, the conversion of forest releases carbon stored in standing biomass and soil carbon. That contributes 39,333 kgCO₂e, or about 63% of the newly harvested timber scenario’s total result.

Using reclaimed timber avoids the need to produce new timber. It keeps an existing material in use.

One pavilion, three timber lifetimes

The assessment compares both options over a 50 year life cycle of a building, as per scientific practice for LCAs.

The newly harvested hardwood, bengkirai, scenario assumes a 20-year material service life, where most of the pavilion timber needs to be replaced due to weathering and degradation. The life cycle model therefore includes the original pavilion and two replacement cycles.

The reclaimed ironwood pavilion, made of reclaimed ulin timber, is modelled as having a 50 year life cycle, lasting for the full study period.

And at the end of life, at 50 years, much of the timber material is still usable. The planks and boards used for flooring and wall cladding will likely be used for firewood, but the posts and beams and larger timber pieces will live on, in another building.

Operational energy, mainly coming from electricity consumption for lighting and the water feature, is identical in both scenarios because the pavilion is the same. It contributes 15,936 kgCO₂e to each total.

The material decision still determines most of the difference:

Newly harvested bengkirai, excluding shared operational energy: 46,270 kgCO₂e
Reclaimed ironwood, excluding shared operational energy: −3,476 kgCO₂e

A longer service life does not make a material impact-free, but it does mean the impact of producing it does not recur as often.

What the numbers mean

The reclaimed scenario has a negative result at the materials stage because the assessment includes the carbon already stored in the recovered timber.

That does not mean the pavilion removes carbon from the atmosphere. It means the reclaimed route receives a credit for keeping existing material in use rather than producing new timber.

The full result is the important number:

12,459 kgCO₂e for reclaimed ironwood
62,205 kgCO₂e for newly harvested bengkirai

We show the negative value because the assumptions should be visible, not hidden, in line with LCA reporting standards.

Eco-Mantra life cycle assessment poster mounted on the pavilion wall, comparing reclaimed ironwood with newly harvested tropical hardwood. — The same comparison, as visitors saw it on the pavilion wall.

What a material label leaves out

“Timber” tells you what something is.

It does not tell you:

where it came from
whether it has already been used
how long it will last
whether it will need replacing
what happened to the land it came from
what will happen to it afterward

Two visitors talking at a reclaimed ironwood table with a planted aquascape set into it, under the slatted pavilion roof in late afternoon light. — A reclaimed ironwood table inside the pavilion.

A life cycle assessment cannot put a number on every consequence of forest conversion. This study does not quantify all biodiversity loss, watershed impacts, community impacts, or indirect effects of new access roads.

It also does not suggest that reclaimed timber is unlimited. There is only so much existing material available, and buying reclaimed wood does not restore a forest that has already been cleared.

It shows something narrower, but useful:

For this pavilion, under these assumptions, reclaimed ironwood has a substantially lower 50-year climate impact than the newly harvested hardwood timber, bengkirai, alternative.

The decision is easier to change early

The message we want to convey through this study is not that reclaimed ironwood will always be the right answer. We have to be careful, as it will not be available forever.

Rather, the message here is that sustainability cannot be inferred from the word “timber” alone.

Where did it come from? How long will it last? Will it need to be replaced? What happens to the timber when the project is dismantled?

Those questions are easiest to answer while a project is still being designed. Once drawings, procurement, and budgets are fixed, changing the material becomes harder, and more expensive.

We measure these decisions early so project teams can act on the results.

Have a material decision you are unsure about?

We help project teams understand what their choices cost, not only to build, but to operate, maintain, replace, and eventually remove.

If you have a project in design and a material decision you cannot settle, send it to us. We will tell you what would need to be measured and where to begin.

Talk to us about your project → See how we work →

Sources

The forest and species research on this page is published work by others. Listed so you can check it.

Mongabay, Where the logging ends in Indonesian Borneo, the forest clearing begins
Mongabay, Indonesia’s dying timber concessions, invaded by oil palms, top deforestation table
Mongabay, Indonesia forest-clearing ban is made permanent, but labeled ‘propaganda’
EIA International, Indonesia’s moratorium on clearing forests and peatlands now permanent, but excludes vast areas
BGCI, Ironwood Eusideroxylon zwageri
EIA International, Ulin: a precious landscape protected by those who live in it
Abdurachman, Susanty and Ngatiman, Analysis of growth and vitality class of Ulin (Eusideroxylon zwageri T. et B.), IOP Conference Series: Earth and Environmental Science, vol. 959, 012071