Page 98 - 2417

Stichnothe et al. Carbon Footprints 2026, 5, 11 Page 3 of 17

that the oil FFB yield of independent farmers is low, representing 42% of the attainable yield . Economic
[22]
constraints and knowledge gaps among smallholders represent significant barriers to productivity. Moreover,
independent smallholders do not have direct access to palm oil mills; frequently, they depend on middlemen
and thus receive lower FFB prices. This leads to below average incomes of independent small holders, as
confirmed by several publications [20-24] .

Xin et al. modeled an increasing land requirement of 19 to 46 million ha oil palm plantations in Indonesia by
2050, with 20% to 50% associated with land use change from peatland and a smaller share from secondary
forests . Despite regulatory measures by the Indonesian government, extensive conversion of peatland and
[25]
other land uses has occurred . Dohong et al. show that over 40% of oil palm plantations located in central
[26]
Kalimantan are situated in deep peat areas .
[27]
GHG emissions from palm oil production on carbon-rich peatland are responsible for a substantial portion
of the country’s total climate impact . That is already confirmed in various papers [28-30] . However, all of them
[5]
used an attributional approach and used different modeling choices for by-products. Lam et al. provide
[28]
spatially explicit calculation of GHG emissions of crude palm oil (CPO) at the level of administrative regions
in Indonesia, considered kernels as by-products, and used economic allocation; the calculated carbon
footprint (CF) was 0.7 to 26 t CO t . Schleicher et al. examined CPO and neglected by-products. They
[29]
-1
2eq
calculated 4 to 30 t CO t . Wang et al. investigated refined palm oil, defined kernel as by-product and
[30]
-1
2eq
used mass allocation in order to allocate GHG emissions, they calculated a CF for refined palm oil of 2.2 t
CO t . The consequence of the modeling choice was not investigated in those papers.
-1
2eq
Understanding the carbon footprint of palm oil, particularly from land-use change (LUC) and through
robust life cycle assessment (LCA) modeling, is paramount for robust decision support for sustainable policy
and mitigating climate change. Therefore, this study focusses on the carbon footprint of palm oil production,
particularly the contribution of direct LUC from forest and/or peatland as well as the influences of modeling
choices regarding by-product substitution.

SYSTEM DESCRIPTION
Extensive research highlights the land-use efficiency of palm oil compared to other vegetable oils; however, a
disadvantage is the poor GHG performance when LUC, particularly from peatland, occurs.

In Figure 1 the considered system boundary for the oil palm plantation is shown. The reported area for oil
palm plantations in national statistics is the productive area, i.e. harvested area marked by dashed lines. The
area required at the nursery stage, the End-of-Life and re-planting is not considered. However, land
productivity can be addressed by varying the yield of FFB ha . Usually, palm oil plantations operated 25-30
-1
years; afterwards the palms are felled, and the trunks remain on the plantation. The harvest starts between
year 3 to 5 and the FFBs are harvested manually and transported to the palm oil mill.

The plantation consists of the managed area and the lagoons, while the oil mill is defined as the mill and the
combined heat and power (CHP)-plant. The palm oil mill effluent (POME) is stored in lagoons and is
afterwards used to irrigate the plantation. It is assumed that fronds remain on the plantation and empty fruit
bunches (EFB) from palm oil mills are returned to the plantation to recycle nutrients and to reduce costs for
fertilizers.

It is well known that the residue management of the palm oil mill has a great influence on the environmental
performance of palm oil production [31,32] . In this study we assume that all residues are utilized; the