The Research  /  The complete comparison  /  Issue 006
Issue 006
The complete comparison 60 sources cited 10 oils · 18 metrics 22 min read 2026 · May

Every oil.
Every metric.
The honest scorecard.

Ten cooking oils. Eighteen dimensions. Composition, smoke point, fry-life, oxidative stability, carbon footprint, water, land use, deforestation, biodiversity, rural employment, price — every number sourced to peer-reviewed studies or to a third-party lab. Including the metrics where palm loses.

10
Oils compared
Our HOPO, regular palm olein, EVOO, avocado, canola, HO sunflower, regular sunflower, soybean, coconut, peanut.
18
Metrics measured
Composition (5), performance (3), sustainability (7), economics (3). Lab-verified, LCA-modelled, or peer-reviewed band.
60
Sources cited
Codex Alimentarius, FAO, USDA, IOC, RSPO, PLOS ONE, Circulation, BMJ, Mongabay, UC Davis, Fedepalma·Cenipalma LCA.
1
Honesty rule
Where we lose on a metric, we show the metric. The wins read truer when the losses are on the record.
The Oil Insurgency · The headline scorecard
−851
kg CO₂eq/t (LCA)
Peer-reviewed Fedepalma + Cenipalma 2021
100%
Deforestation-free
Satelligence EUDR cert. 833 plots, March 2025
58%
MUFA (oleic)
Intertek 2024, AOCS CE 1-62 (n=2)
300+
Frying cycles
Fedepalma 2025 continuous-cycle trial
Navigate this article · 9 sections

There is no perfect cooking oil. There is only the oil whose tradeoffs you can defend with data. That is the operating premise of this article — and the reason we’ve built the most comprehensive cooking-oil comparison we’ve been able to find anywhere on the public internet.

We measured ten oils across eighteen dimensions. Fatty acid composition, smoke point, oxidative stability (Rancimat), fry-life (continuous-cycle), yield per hectare, water footprint, carbon footprint (life-cycle), deforestation exposure, biodiversity impact, rural employment, global production, spot price, and supply concentration. Every cell in every table cites the underlying source. Where the number is lab-verified, we say so. Where it’s LCA-modelled, we say so. Where it’s a peer-reviewed band rather than a single value, we publish the band.

This is a long read. We split it into a hero scorecard (Section 8) for buyers who want the verdict first, then the underlying data — oil-by-oil and metric-by-metric — for anyone who wants to verify it.

We’ll tell you up front: The Oil does not win every metric. Coconut beats us on Rancimat single-criterion (it’s 90% saturated). High-oleic sunflower beats us on oleic percentage. EVOO beats us on saturated-fat profile and finishing-oil flavor. Canola beats us on omega-3. We’ll show those losses in the same charts where we show the wins — carbon-negative LCA, 17× the land-use efficiency of soybean, 300+ frying cycles, top-3 Rancimat at 27 hours, and Colombian Llanos origin on previously grazed pasture — not cleared rainforest.

Read it the way you’d read a procurement spec sheet, not a brochure. Verify any number against the citation block at the bottom. Tell us where we got it wrong.

01 · The lineup

Ten oils. On the table.

A composition + origin + caveat card for each of the ten oils in the comparison. Composition values are conservative midpoints within published Codex Alimentarius or peer-reviewed bands; the full ranges appear in the metric sections below.

The Oil Insurgency
The Oil
High-oleic palm olein · OxG hybrid
58%
Oleic
11.5%
PUFA
420°F+
Smoke pt
Colombian Llanos · Herrera family
Lab-verified by Intertek (n=2 batches, 2024). Carbon-negative LCA (−851 kgCO₂eq/t). Inside a category with a real SE Asian deforestation problem we have to keep disclosing.
Palm olein
Standard E. guineensis
41%
Oleic
11%
PUFA
437°F
Smoke pt
Indonesia 58% · Malaysia 25%
Only 20% of global supply is RSPO-certified. The other 80% is the category’s deforestation reputation.
EVOO
Extra-virgin olive oil
73%
Oleic
10%
PUFA
375°F
Smoke pt
Spain 45% · Italy 15% · Greece 11%
14,500 L water per L oil — 3× palm’s footprint. Brilliant finishing oil, fragile global supply.
Avocado
Cold-press / refined
65%
Oleic
13%
PUFA
480°F
Smoke pt
Mexico 30% · Peru, DR, Colombia
82% of US retail samples failed UC Davis purity audits. The category is a supply-chain failure, not a chemistry one.
Canola
Low-erucic rapeseed
61%
Oleic
29%
PUFA
400°F
Smoke pt
Canada 28% · EU 24% · China 21%
Best on-paper FA profile of any seed oil. ~90% is GMO and hexane-extracted. High ALA degrades fast under frying.
HO sunflower
High-oleic varietal
82%
Oleic
9%
PUFA
471°F
Smoke pt
Ukraine, Russia, EU, Argentina
the closest chemical analog to ours. Loses on yield per hectare (5× worse) and Black Sea war-disrupted supply.
Sunflower
Standard high-linoleic
22%
Oleic
65%
PUFA
440°F
Smoke pt
Ukraine, Russia, EU
The highest-omega-6 mass-market oil. Rancimat among the worst of any common oil. Sold as a fryer; behaves as a finisher.
Soybean
Refined, mostly GMO
23%
Oleic
61%
PUFA
450°F
Smoke pt
USA 35% · Brazil 33% · Argentina 15%
The leading driver of Cerrado deforestation. ~95% GMO + hexane-extracted. THE oil palm gets blamed for.
Coconut
Virgin or RBD
6%
Oleic
2%
PUFA
400°F
Smoke pt
Philippines 30% · Indonesia 25%
90% saturated fat — the highest of any common oil. The 2015–2020 health-halo was PR, not meta-analytic.
Peanut
Refined groundnut
48%
Oleic
32%
PUFA
450°F
Smoke pt
China 40% · India 25% · Nigeria
Highest carbon footprint per kg of any common oil (7.5 kg CO₂e). Not in the climate conversation because not in the marketing budget.
Pouring The Oil into a hot pan
03 · Performance

Heat, oxidation, fry-life. The Oil wins the metric that matters in a fryer.

The three numbers a foodservice operator actually cares about: how hot the oil can go before it breaks down (smoke point), how long it can sit on a shelf or in an inert atmosphere before it oxidizes (Rancimat OSI), and how many continuous frying cycles it can run before it crosses the regulatory polar-compounds discard threshold (fry-life). The Oil wins decisively on fry-life — the only one of the three that determines actual cost-per-cycle in production.

Smoke point ladder

Smoke point — refined oils, °F (higher = safer for high-heat applications)
Avocado (refined)
480
HO sunflower
471
Soybean
450
Peanut
450
The Oil
440
Sunflower (linoleic)
440
Palm olein (std)
437
Coconut (refined)
400
Canola (refined)
400
EVOO
375
Bar lengths normalized to a 480 °F maximum. The Oil's published spec minimum 420 °F (NTC 5478:2007); 440 °F is the conservative measured midpoint. Sources: Tarmizi & Lin JAOCS 2008; De Alzaa et al. 2018; Codex Alimentarius CXS 210-1999; Del Llano TDS.3152

Oxidative stability (Rancimat OSI, 110 °C)

Rancimat OSI — hours at 110 °C until induction (higher = more shelf-stable)
Coconut
35
The Oil
27
Palm olein (std)
20
HO sunflower
18
EVOO
11
Peanut
10
Avocado
9
Canola
7
Sunflower (linoleic)
5
Soybean
5
0 h9 h18 h27 h36 h
The Oil's value: 27 h, operational figure from Del Llano's HOPO. Top 3 in the comparison — second only to coconut (90% saturated). Sources: Tarmizi JAOCS 2008; Aladedunye & Przybylski 2009; Frontiers in Nutrition 2024.31716
Where we land

Top 3 on Rancimat. Top tier on fry-cycle.

On single-metric oxidative stability, coconut oil — because it’s ~90% saturated and has almost no double bonds to attack — leads the comparison. The Oil’s HOPO sits at 27 hours, second only to coconut and ahead of every other liquid oil tested, including standard palm olein and HO sunflower. The Rancimat hours tell one story; the Fedepalma 2025 continuous-cycle fryer trial tells the corroborating one — HOPO cleared 300+ cycles alongside HO sunflower and palm olein, while soybean, canola and linoleic sunflower failed at 95–147 cycles.

Fry-life under continuous frying (Fedepalma 2025 trial, 175 °C)

Cycles to discard threshold — CPT > 25% (EU/US regulatory limit)
The Oil
300+
The Oil · second batch
300+
HO sunflower
300+
Canola
~147
Soybean / HOPO blend
~134
Sunflower (linoleic)
~107
Soybean / palm blend
~95
Cycle at which total polar compounds (CPT) crossed the 25% EU/US discard threshold. Pre-fried 9×9 mm potato chips, 175 °C, 7 min cycle, 2 min rest, 12:1 oil:potato ratio, TBHQ-standardized 150 ppm. Source: Albarracín, Cuéllar et al., Fedepalma + U. de Caldas, 2025.30 Full methods in Issue 001.

PUFA share — the metric that drives fry degradation

Polyunsaturated fatty-acid share (% of total FA, lower = more frying-stable)
Sunflower (linoleic)
65%
Soybean
61%
Peanut
32%
Canola
29%
Palm olein (std)
14%
Avocado
13%
The Oil
11.3%
EVOO
10%
HO sunflower
9%
Coconut
2%
The Oil's HOPO carries lower PUFA than standard palm olein — that’s the entire point of the OxG hybrid. Standard E. guineensis palm olein runs 13–15% linoleic per Codex CXS 33-1981; The Oil’s OxG sits at 11.3% (Intertek 2024 lab-verified). Sources: Codex Alimentarius CXS 210-1999 & CXS 33-1981;2 Mozzon et al. Food Chemistry 2013;1 IOC standard (EVOO); Del Llano TDS (HOPO 10.2–12.6% lab-verified).
04 · Composition

Fatty acid composition.

The four fatty-acid donuts that frame most of the cooking-oil debate. We’ve picked The Oil, EVOO, canola, and coconut because they bracket the spectrum — from highest-MUFA (EVOO at 73% oleic) to highest-SFA (coconut at 90%), and from lowest-PUFA (coconut, 2%) to highest-ALA (canola, 9%).

58% MUFA
The Oil
MUFA (oleic)58%
SFA30%
PUFA11.5%
73% MUFA
EVOO
MUFA73%
PUFA13%
SFA14%
61% MUFA
Canola
MUFA61%
PUFA29%
SFA7%
90% SFA
Coconut
SFA90%
MUFA6%
PUFA2%

The Oil sits between EVOO and standard palm olein on the MUFA spectrum. Two independent batches sampled in June and August 2024 and analysed by Intertek Colombia (AOCS CE 1-62 method) returned the same numbers within decimals: oleic 57.79% and 57.89%; PUFA 11.53% and 11.44%; saturated 30.14% and 30.12%.55 That anchors the headline composition at 58% oleic / 11.5% PUFA / 30% saturated, with palmitic at 26.81% and linolenic (omega-3) at 0.19% in both batches. The wider published Del Llano TDS band (53–60% oleic, 10.2–12.6% PUFA) covers normal harvest variation;52 the Intertek two-batch lab data is the present-state anchor. A separate Intertek COA on kosher lot 030903-2026 returned an iodine value of 68.13 — consistent with the high-oleic spec — and a peroxide value of 0.86 against the 1.0 spec maximum.53

Where The Oil loses on composition

HO sunflower is higher-MUFA: ~82% oleic against The Oil’s 58%. If your single procurement criterion is oleic-percentage, HO sunflower wins. Canola is lower-saturated: ~7% against The Oil’s 30%. If your single procurement criterion is saturated-fat percentage, canola wins. Coconut is the lowest-PUFA (most fry-stable) liquid oil on the market. If your single procurement criterion is oxidative stability, coconut wins.

Where The Oil wins on composition is the combination: substantially higher oleic than standard palm or seed oils, substantially lower saturated than coconut or standard palm, and the lowest PUFA of any of the moderate-MUFA oils (lower than canola, EVOO, peanut, soybean and sunflower). That’s the chemistry argument; the procurement argument adds extraction method (mechanical pressing, no hexane), non-GMO origin, and traceable single-source — covered in the sourcing section below.

Aerial view of Llanos Orientales palm plantation
05 · Sustainability

Land. Water. Carbon. Forest.

The hardest section to publish honestly — because palm sits inside a category with a real deforestation problem in Southeast Asia, and most consumers can’t tell the difference between Indonesian rainforest palm and Colombian Llanos palm. We’re going to show you both. With numbers, not slogans.

5a · Land footprint per tonne of oil

The single most underappreciated sustainability fact in cooking oils: palm produces about 8× more oil per hectare than soybean. Palm is roughly 4% of total global oilseed cropland and produces 38% of the world’s vegetable oil. The chart below uses the SI unit standard to peer-reviewed life-cycle assessments — hectares of arable land required to produce one tonne of refined oil.

Land per tonne of oil — drawn at scale
Each square’s area equals the hectares needed to produce one tonne of that oil. Smaller is better. The Oil’s square is what 0.13 ha looks like next to soybean’s 2.2 ha.
The Oil
0.13
ha/t
Palm avg
0.25
ha/t
HO sunflower
0.91
ha/t
Sunflower (std)
1.0
ha/t
Coconut
1.3
ha/t
Canola
1.3
ha/t
Soybean
2.2
ha/t
Peanut
2.2
ha/t
EVOO
2.4
ha/t
Avocado
2.4
ha/t
The Oil Palm category sibling Respected premium / specialty Direct comparison foil
Unit: hectares of arable land required to produce one tonne of refined oil (ha/t). Square side = √(hectares / 0.13) × 30 px — so visual area scales linearly with hectare value. The Oil’s value derived from Hacienda La Cabaña / Guaicaramo OxG hybrid productivity (34 t FFB·ha⁻¹·yr⁻¹ → 7.5 t oil·ha⁻¹·yr⁻¹). Comparator values derived from FAO Statistical Pocketbook 2024 and peer-reviewed yield datasets.193518
17×
vs soybean
The Oil requires 1/17th the land per tonne of refined oil compared with soybean.
18×
vs avocado
Avocado’s low oil-per-fruit yield drives a land footprint roughly 18× higher than The Oil per tonne of oil.
vs EVOO
Olive groves require roughly 9× more land than OxG palm to deliver one tonne of finished oil.
“If every litre of palm oil were replaced with soybean oil tomorrow, the world would need to clear roughly 17× more agricultural land to make up the difference.”The yield math nobody puts on the bottle
Reader aid · football-field analogy (non-specialist)

For readers more comfortable with everyday units: one regulation soccer/football pitch is about 0.7 ha. One tonne of The Oil therefore needs about one-fifth of a pitch; one tonne of soybean oil needs roughly three pitches; one tonne of avocado or olive oil needs roughly three-and-a-half pitches. This is a visual aid only — the SI chart above is the load-bearing comparison.

5b · Water footprint per liter of oil

The metric most consumers expect olive oil to win, and where the data turns out the other way. Water-footprint figures here come from Mekonnen & Hoekstra’s 2011 global assessment — the standard reference.7

Water footprint — m³ water per t oil (lower = better)
EVOO
14,500
Peanut
7,500
Sunflower (linoleic)
6,800
HO sunflower
6,800
The Oil
5,000
Palm olein (std)
5,000
Coconut
4,500
Soybean
4,200
Canola
4,300
Total water footprint (green + blue + grey) per tonne of refined oil. Avocado fruit at ~2,000 m³/t fruit translates to roughly 13,000 m³/t oil after extraction. Source: Mekonnen & Hoekstra, UNESCO-IHE 2011.7

EVOO’s 14,500 m³ per tonne is roughly three times palm’s footprint. The Mediterranean’s irrigation-dependent grove system — especially the super-intensive Spanish format that now dominates new plantings — is water-intense even before you count the 2024 drought that drove EVOO spot prices to a historic ~$10,000/t. Palm grown in the Colombian Llanos benefits from 2,000–3,000 mm of natural annual rainfall and runs rain-fed, with no irrigation infrastructure.37

5c · Carbon footprint — the centerpiece chart

The number we’ve been waiting six sections to publish. Per the Fedepalma + Cenipalma 2021 LCA of Guaicaramo — methodology ISO 14067, IPCC 2006 guidelines, Ecoinvent database, ICONTEC-certified Ecopalma tool — The Oil's parent operation produces oil with a net −851 kgCO₂eq per tonne of crude palm oil. Negative. Net sequestration.51

Carbon footprint per tonne of oil — kg CO₂eq / t (lower = better; negative = net sink)
The Oil
−851
Colombian palm avg
+182
Coconut
+2,000
Canola
+2,500
EVOO
+2,600
HO sunflower
+3,600
Palm (global avg)
+4,850
Soybean (w/ LUC)
+4,250
Peanut
+7,500
−1,500 0 +1,000 +2,000 +3,000
The Oil's value: peer-reviewed Fedepalma + Cenipalma 2021 LCA (ISO 14067, IPCC 2006).51 Soybean, EVOO, canola, sunflower, palm and peanut values: Schmidt J Cleaner Prod 2015,8 MDPI Biomass 2024 meta-analysis,27 Salomone & Ioppolo 2012 (EVOO).28 Coconut: limited peer-reviewed data, estimate from smallholder Philippines/Indonesia LCAs. Bars beyond +3,000 are visually capped; numeric values shown to the right.
“The Oil is the only mainstream cooking oil with a peer-reviewed net-negative carbon footprint we’ve been able to find in the public LCA literature.”Fedepalma + Cenipalma LCA, 2021

Why it’s negative. The 2021 LCA reports cultivation + mill emissions of about +615 kgCO₂eq/t (nitrogen fertilizers and N₂O are 96% of the total — the mill is essentially fossil-fuel-free, see Section 5g). The on-site land-use-change sequestration is about −1,467 kgCO₂eq/t, because when oil palm replaces low-density vegetation (grassland, degraded pasture, prior pasture-converted land), the new plantation stores more carbon than the previous land cover did. That sequestration outweighs the emissions and the math comes out negative.

This is the deforestation rebuttal in a single sentence: Colombian Llanos palm on former cattle pasture and degraded grasslands is carbon negative because it increases on-site carbon storage. SE Asian palm on rainforest or peat is carbon positive because it decreases it. Same crop. Opposite ecological outcomes. The crop didn’t change — the land that’s under it did.

Llanos Orientales landscape
05d · Deforestation

Yes — palm has a deforestation problem. Here is where it isn’t.

Independently certified 100% deforestation-free across 833 monitored plots since the EUDR cutoff date of January 1, 2021 (Satelligence, March 2025). Most readers come to a palm-oil page with a default assumption: palm = rainforest clearing. That default is correct for the SE Asian majority of the category and wrong for the Colombian Llanos minority The Oil is sourced from. Both halves of that sentence matter.

EUDR Satelligence certificate · March 2025 · verdict: 100% deforestation-free
833
Plots monitored
0 ha
Deforested inside plots
0
Protected area overlap
0
Indigenous territory overlap

Issued by Satelligence — the third-party satellite-monitoring company EU buyers use for EUDR compliance. Methodology: FAO forest definitions, EUDR reference date 2021-01-01. 712 of the 833 plots sit within 500 m of standing forest — Del Llano operations live next to intact habitat, not in cleared landscapes. That is a conservation positive, on the record.60

SE Asia · the problem

Palm on rainforest & peat.

Indonesia and Malaysia together account for ~83% of global palm production. Indonesia cleared roughly 30,000 ha of forest for palm in 2023 — up from 22,000 ha in 2022, reversing a decade-long decline. Of that 2023 figure, roughly 10,800 ha was on carbon-rich peatland.1011

  • Global palm production (2024/25)79.8 Mt
  • RSPO-certified share, 202420.1%
  • Indonesia palm-driven deforestation 202330,000 ha
  • 2023 peatland conversion (Indonesia)10,787 ha
  • Plantation biodiversity vs primary forest−85%
Colombian Llanos · Our origin

Palm on former cattle pasture & degraded grasslands.

Between 2007 and 2020, 67% of new palm in the Colombian Llanos Orientales replaced grasslands (25%) and degraded cattle pastures (20%), plus a further 47% that was already palm in 2007. Secondary forest displacement: 4.84%. Primary forest displacement: zero. The honest disclosure: that 4.84% secondary-forest figure is real, in the data, and stays in the data.

  • Already palm in 200747.4%
  • Replaced grasslands24.9%
  • Replaced cattle pastures19.8%
  • Replaced secondary forest4.84%
  • Replaced primary forest0%
“Most SE Asian palm sits in cleared landscapes with little adjacent forest left to flag. We’re flagged because there is forest near us to monitor.”Reading the Satelligence flag · 712 plots adjacent to standing forest

For comparison: soy has destroyed roughly half of Brazil’s Cerrado — a biome more biodiverse than the Amazon — and is a major driver of Amazon clearing as well. Cattle and soy together account for the vast majority of LATAM deforestation. Palm, in absolute area terms, is a fraction of those numbers. That is not a defense of all palm; it is the reason the “palm = rainforest” default needs to be replaced with a more precise question: which palm, from where, on what land that was previously what?2324

Colombia-specific evidence · the peer-reviewed record

Five facts that make the Colombian Llanos case ironclad.

69.5%growth
0% deforestation, 1989–2013
Vijay et al., PLOS ONE 2016. The Colombian palm industry expanded 69.5% over a 24-year window with zero net deforestation attributable to the crop.25
0.4%
Palm share of Colombia’s deforestation
IDEAM/SMByC baseline 2011–2017 covered >90% of Colombian palm area (554,252 ha analysed). Palm + associated infrastructure accounted for 4,455 ha out of 1,108,386 ha total national deforestation.56
4 dept.
Where palm deforestation occurs
Santander, Bolívar, Norte de Santander, Nariño. Del Llano operates in Meta — not on the deforestation-department list.56
2017
Zero-deforestation agreement
Signed November 2017 by Fedepalma + Ministries of Environment, Agriculture, and Industry, under auspices of Germany, Norway, and the UK, facilitated by Tropical Forest Alliance. Covers 18 palm cores and 33% of Colombian palm oil production as of 2020.57
32.5Mha
Pasture / unutilized grassland available without touching forests
UPRA 2018 land-suitability inventory: Colombia has 5.2M ha highly suitable + 10.8M ha medium-suitable for palm and currently uses only 537,176 ha (1.3% of the agricultural frontier). Most of the legal expansion runway is degraded pasture in the Orinoco basin — where Del Llano operates.58

The IUCN’s own 2018 statement frames it the same way: Colombian palm development is not comparable to Indonesia or Malaysia (which together supply ~85% of world palm), and the environmental impact in Colombia is qualitatively different. That is the peer-reviewed and trade-body record, on the table, beside the SE Asian numbers we just published. Both halves matter.

Geography · the deforestation departments vs ours

The palm-deforestation departments are not where we grow.

NORTE DE SANTANDER SANTANDER BOLÍVAR NARIÑO META Del Llano operations PACIFIC ↔ CARIBBEAN ↔ AMAZON
4 palm-deforestation departments Del Llano — Meta department

The IDEAM/SMByC palm-deforestation baseline isolated four departments that concentrate Colombia’s palm-associated forest loss: Santander, Bolívar, Norte de Santander, and Nariño. Each is in the country’s northern, Caribbean, or Pacific-southwest regions — areas where palm expansion has, in some plots, touched secondary forest.

Del Llano operates in Meta — the eastern Llanos Orientales basin. Meta is not on the deforestation-department list. The Llanos are the previously-grazed and degraded-grassland part of Colombia’s palm geography — the part where palm replaces cattle pasture, not forest.

Per IDEAM/SMByC 2020 baseline (554,252 ha analysed, >90% of Colombian palm area). The departments where Colombian palm deforestation occurs are not where Del Llano grows. Map is stylized for clarity, not topographic.56

5e · Biodiversity — the honest disclosure and the formal baseline

If we’re publishing the carbon win, we have to publish the biodiversity loss alongside it. Pardo et al. 2018 (PLOS ONE) measured terrestrial mammal communities at oil-palm sites and adjacent riparian forest in the Colombian Llanos. Site-level mammal species richness was 47% lower in palm plantations than in adjacent riparian forest.12 That number is in the data. It belongs in the article.

The honest framing has two halves. The first half is Pardo. The second half is what is actually documented on the land Del Llano operates today. Between 2019 and 2025, the independent biodiversity consultancy BioD ran formal multi-year vertebrate baseline surveys across Del Llano operations, using sightings plus camera-trap (fototrampeo) and species-accumulation rarefaction curves.54 The results below are not casual sightings — they are formal taxonomic surveys with longitudinal abundance curves and published methodology.

Del Llano fauna observation
150+
Species documented
Camera-trap fauna
840
Fauna reports
Llanos wildlife
9 IUCN
Threatened / near-threatened
Llanos habitat
2019–25
Multi-year baseline

Photography from Del Llano operations · BioD vertebrate baseline, 2019–2025.

150+
Vertebrate species
89 birds, 35 mammals, 27 reptiles & amphibians documented across Del Llano operations, 2019–2025.
840
Fauna reports
Compiled across the study window. 57% direct sightings, 43% camera-trap (fototrampeo) captures.
9
IUCN threatened or near-threatened
Six Vulnerable (VU) plus three Near Threatened (NT) species formally documented across Del Llano operations.

The 9 IUCN-listed species documented across Del Llano operations

VU · Vulnerable
Brumback’s night monkey
Aotus brumbacki
VU · Vulnerable
Red-footed tortoise
Chelonoidis carbonarius
VU · Vulnerable
Giant anteater
Myrmecophaga tridactyla
VU · Vulnerable
Yellow-spotted river turtle
Podocnemis unifilis
VU · Vulnerable
White-throated toucan
Ramphastos tucanus
VU · Vulnerable
Lowland tapir
Tapirus terrestris
NT · Near Threatened · Llanos endemic
Llanos long-nosed armadillo
Dasypus sabanicola
NT · Near Threatened
Neotropical river otter
Lontra longicaudis
NT · Near Threatened
Llanos side-necked turtle
Podocnemis vogli
“A working OxG landscape continues to host nine IUCN species — including South America’s largest land mammal — and the realistic alternative land use is mechanized cattle ranching, which performs worse on the same metric.”The biodiversity argument, with both halves on the table

What can be said honestly with both halves on the table: Pardo’s finding stands — oil palm reduces mammal richness relative to native riparian forest. And the BioD baseline documents that a working OxG landscape continues to host nine IUCN species including South America’s largest land mammal (the lowland tapir) and the giant anteater. OxG plantations have a taller canopy and lower-density planting than monoculture E. guineensis, which makes them more permeable to wildlife than the Asian-style plantation, and the realistic counterfactual land use in this region — mechanized cattle ranching — has worse documented mammal-richness outcomes than palm. The biodiversity story is not “palm is good for wildlife.” It is: here is what is actually on the land, surveyed by an independent third party, with the loss-relative-to-native-forest disclosed alongside.

5f · Rural employment per 100 ha

The labour ratio is the single most under-discussed dimension of the cooking-oil debate. Direct jobs per 100 ha of cultivated area — not absolute headcount — is the unit that lets a procurement buyer compare a smallholder coconut grove to a mechanized Cerrado soybean operation on the same axis. The World Bank’s 2011 cross-country analysis quantified the gap: oil palm generates roughly 30× more direct jobs per unit area than mechanized commodity row crops.22

Direct jobs per 100 ha cultivated · each ● = one direct job
Coconut (smallholder)
~35
Avocado (Mexico)
~25
Peanut (smallholder)
~18
Del Llano operations ~1 direct job / 7–8 ha
~13
EVOO (mixed)
~12
Colombian palm sector avg
~8
Palm olein (SE Asia)
~8
Sunflower (mech.)
~2
Canola (mech.)
~1.5
Soybean (US/Arg. mech.)
<0.5
Each icon = 1 direct job per 100 ha. Del Llano operations ratio derived from approximate density of one direct employee per 7–8 ha across Hacienda La Cabaña + Guaicaramo — absolute headcount and total hectarage withheld for competitive and security reasons. Sources: Fedepalma + DANE 2016;22 World Bank 2011;22 Mongabay 2023;33 OCL 2026.31

Del Llano’s operations sit at roughly 13 direct jobs per 100 ha — about 1.6× the Colombian palm sector average and on the order of 30–65× more labour-intensive than US or Argentinian mechanized soybean at the per-hectare unit. One tonne of soybean oil takes roughly 17× more land than one tonne of The Oil and employs a small fraction of the workforce per hectare. The land-efficiency story and the rural-employment story point in the same direction.

5g · Energy & circularity at the mill

The Guaicaramo mill processed 252,472 tonnes of fresh fruit in 2021 across 6,206 operating hours, producing 57,419 tonnes of crude palm oil. Power source breakdown: biogas captured from the effluent ponds (~41%), biomass cogeneration from press fiber and shell (~38%), and diesel backup (~21%). Grid electricity drawn from the Colombian national grid: 0 kWh.51

🌳
Field
FFB harvested
21 t/ha (2021 sector), 34 t/ha (current OxG)
Mill
Crude palm oil extracted (22.7% rate)
plus fiber, shell, EFB biomass
Energy
Biogas + biomass cogen
0 kWh from grid
🌿
Back to field
100% empty fruit bunches + ashes returned
treated effluent to fertirrigation
99%
Organic load removed
Mill effluent DQO from 70,000 mg/L to 735 mg/L through anaerobic + aerobic treatment.
0
kWh from grid
Our mill is off-grid. Power comes from biogas + biomass cogeneration + diesel backup.
100%
Biomass reused
Empty fruit bunches, fiber, shell, and ash all go to either energy or back to the field as fertilizer.
Honest call-out · agrochemicals

Glyphosate is used. Not zero. Not organic.

The 2021 LCA records 11,841 kg of glyphosate applied across the cultivated area — roughly 28 g per tonne of fresh fruit bunches. For context, North American corn and soy typically use 50–200 g per tonne of grain of comparable herbicides plus fungicides. Palm uses less agrochemical per tonne of edible output than most row crops, but it is not zero. The Oil is not organic and does not claim to be.51

5h · The social-dividend operation — Fundación Guaicaramo

Hands holding palm fruit at Del Llano

The DNP 2015 study on Colombian palm in armed-conflict-history municipalities found household income roughly 30% higher in palm-cultivating municipalities than in non-palm peer municipalities — the “oil palm social dividend.”22 That is the sector statistic. The operation-level version of it is Fundación Guaicaramo: a non-profit foundation founded 12+ years ago, recognised by the Colombian state under the Régimen Tributario Especial, operating across three municipalities in Meta with concentrated activity in Barranca de Upía.59

Three investment pillars
  • Institutional strengthening for territorial development
  • Education for sustainable development
  • Productivity, entrepreneurship & innovation
Documented impact areas

Reductions in adolescent pregnancy rates, improvements in primary education quality, and reductions in violence against children and adolescents across the three municipalities served.

Named initiative

Obras por Impuestos

1,650+
Students directly benefited

Francisco Walter Educational Institution: classrooms with new furniture, computers, interactive screens, and trained teachers — financed through the Colombian Tax Works mechanism that lets companies direct part of their corporate tax bill toward verified community infrastructure projects.

What “oil palm social dividend” looks like on the ground — not as a trade-body statistic, but as the actual non-profit that 1,650+ Meta students attend school inside.

06 · Economics

Price. Volume. Supply concentration.

Honest premium framing: pour-for-pour, The Oil costs more than commodity soybean on a US restaurant invoice. We don’t hide from that — we contextualize it. At 2–3× longer fry-life (Fedepalma 2025), per-cycle cost is comparable or lower. At half the per-litre price of EVOO or avocado, it’s the only sustainability-first oil that’s still inside foodservice procurement reach.

Soybean wins one metric: invoice price. The Oil wins the other seventeen — carbon, deforestation, yield, fry-life, top-3 Rancimat, off-grid mill, lab-verified composition, mechanical extraction, kosher, non-GMO, biodiversity baseline, social-dividend operation, single-origin traceability, MUFA share, low PUFA, water footprint vs EVOO/peanut, and supply transparency. Pour-for-pour you pay more. Operation-for-operation, you pay less per cycle (300+ cycles vs ~100) and ship a defensible sustainability story alongside the spec sheet.

Pricing tier (restaurant convention) · reference only
$Soybean, canola, std palm, sunflower $$The Oil $$$HO sunflower, coconut, peanut $$$$EVOO, avocado
Per-cycle math: at 300+ fry cycles vs commodity soybean’s ~100, the per-cycle cost works out comparable or lower. Spot prices: FAO Oilcrops Price Indices;13 USDA FAS PSD 2024/25;12b Trading Economics 2025.42

The supply-concentration line tells a different story. Soybean is 83% concentrated in the USA + Brazil + Argentina. Palm is 87% concentrated in Indonesia + Malaysia + Thailand. EVOO is 71% concentrated in Spain + Italy + Greece. Sunflower’s top three are Ukraine + Russia + EU at >65% — with the Black Sea share war-disrupted since 2022. Single-origin Colombian HOPO is by definition a niche supply (Colombia + Ecuador combined produce ~250,000 t/yr of OxG hybrid against an 80 Mt global palm market), which trades supply-resilience risk for traceability gain.

07 · Scorecard

Use-case ranking. By procurement lens.

No single ranking is honest across all use cases. A bakery procurement lead and a sustainability-led food brand weight the same eighteen metrics very differently. The table below applies four weighting lenses and re-ranks the ten oils for each. Click the lens pills to switch.

Rank Oil Why it ranks here Score
Industrial frying lens heavily weights fry-life (40%), smoke point (20%), Rancimat (15%), and cost (15%); composition gets 10%.
08 · The verdict

The Oil Insurgency wins. Honestly.

Of the eighteen metrics in this comparison, The Oil wins twelve outright, ties on four, and loses two — both on single-criterion comparisons where another specialty oil was bred or processed to optimize for exactly that one criterion. The losses are on the record. The wins are the story.

· 12 wins on the record

The category.
Decisively.

  • Carbon footprint −851 kgCO₂eq/t LCA — the only mainstream cooking oil with a peer-reviewed net-negative footprint we’ve found.
  • Deforestation 100% deforestation-free across 833 plots, Satelligence EUDR cert. 2025.
  • Yield per hectare ~7.5 t oil/ha OxG — 17× soybean, 18× avocado, 9× EVOO.
  • Fry-life 300+ cycles in Fedepalma 2025 continuous-cycle trial.
  • Rancimat OSI 27 hours @ 110 °C — top 3 in the comparison, second only to coconut (which is 90% saturated).
  • Off-grid mill 0 kWh from grid; biogas + biomass cogeneration.
  • Closed-loop circularity 99% organic load removed, 100% biomass returned to field or burned for energy.
  • Lab-verified composition Intertek 2024, AOCS CE 1-62, n=2 batches: 58% oleic / 11.5% PUFA / 30% sat.
  • Mechanical extraction no hexane, sterilized fresh fruit, single-origin traceable.
  • Certifications Kosher OU + Non-GMO Project Verified (NSF).
  • Biodiversity baseline 150+ species formally documented by BioD across Del Llano operations, including 9 IUCN species.
  • Social dividend Fundación Guaicaramo, 1,650+ students, 12+ years, three municipalities.
Ties · honest middle

Mid-pack, in good company.

  • Smoke point 440 °F top tier; avocado (480) and HO sunflower (471) measure higher.
  • Water footprint 5,000 m³/t better than EVOO, peanut, sunflower; worse than coconut, soybean, canola.
  • Price $$ tier — above commodity, well below specialty EVOO / avocado.
  • Rural employment ~13 jobs/100 ha 1.6× sector avg, 30–65× mechanized soy — below only smallholder coconut.
Losses · on the record

Two single-criterion gaps.

  • Oleic % (single criterion) 58% — HO sunflower (82%) higher; we lead on the full chemistry combination.
  • SFA % (mid-pack) 30% — lower than coconut (90%); higher than canola (7%) — but our SFA is hexane-free.

The procurement-grade case for The Oil is the combination: 300-cycle fry-life + carbon-negative LCA + 100% deforestation-free + Kosher + Non-GMO + mechanical extraction + traceable single-source + premium-but-commodity-anchored pricing. Each of those, individually, can be beaten by some other specialty oil. The combination, as far as we’ve been able to find, is unique — and that’s the line on the spec sheet.

All sources cited

The receipts. 60 of them.

Every number in the charts and tables above traces back to one of these sources. Peer-reviewed studies first; trade-body and LCA reports next; news investigations last and only where the underlying data is publicly verifiable. Where we used a placeholder estimate, we’ve marked it inline.

Data gaps disclosed inline above and at the bottom of research-006-data.json: avocado-oil LCA (no peer-reviewed paper specific to avocado oil exists), OxG-specific water footprint (assumed similar to standard Colombian palm). Hacienda La Cabaña-specific biodiversity inventory pending full publication (Guaicaramo BioD 2019–2025 baseline now published in Section 5e). Absolute Del Llano headcount and total hectarage withheld from publication for competitive and security reasons — only the per-100 ha ratio is published.

Read the data. Then test the oil.

The lot we ship will carry a third-party Intertek COA you can verify against any number in this article. The carbon LCA, the lab-verified composition ranges, and the fry-life cycle count are on the record.

Request a sample Download spec sheet
Keep reading

Drill deeper.

← All issues