LIT_109: Starck et al. (2024) — excreta-P recycling = 7-34% of French food-P self-sufficiency (P-only)

LIT_109: Starck et al. (2024) — excreta-P recycling = 7-34% of French food-P self-sufficiency (P-only)

Source

doi:10.1007/s10705-024-10367-4 — original publication (opens in a new tab; the file is not redistributed)

Summary

A national phosphorus (P) mass balance of the French sanitation system, built from operational data on ~20,000 wastewater treatment plants (WWTPs) over two decades (Starck, Fardet & Esculier, LEESU / Ecole des Ponts). For Neobiome its value is the metric it quantifies, not its French numbers: it puts a defensible figure on how much of a nation’s food-phosphorus need can be met by recycling human-excreta P. The headline finding is that recycling all P in French excretions could supply 7 to 34% of the P in domestic food supply without changing the food system, a food-self-sufficiency (P-security / phosphate-rock-independence) framing. The 7% reflects the current animal-heavy French diet; the 34% is the ceiling if the recovered P is prioritised to high-efficiency plant-based crops for human consumption.

This is a transferable proxy (France, not NZ; same status as LIT_060) and it is phosphorus only — it carries no nitrogen figure. It is the peer-reviewed precedent for the excreta-nutrient-offset metric the NI model uses, and it is precise about a distinction the model must not blur: the 7-34% food-supply metric is not the same as the ~15% direct fertiliser-consumption offset.

Key claims

- claim: "HEADLINE (food-P self-sufficiency, P-only): recycling all P in human excreta could supply 7 to 34% of the P in French DOMESTIC FOOD SUPPLY. Verbatim (abstract): 'Our results suggest that recycling all P in excretions could help supply 7 to 34% of the P demand in French food supply, without changing the current food system.' The 7% is the current animal-heavy French diet; 34% is the ceiling when excreta-P is prioritised to plant-based production for human consumption. Verbatim (Results §3.3): 'the potential contribution to food domestic supply is only about 7% but it could go up to 34% if plant-based products for human consumption are prioritized.' This is a food-self-sufficiency metric (P in excreta vs the P footprint of domestic food supply), NOT a generic fertiliser-demand offset. [verified verbatim against the raw]"
  source_location: "Abstract; Results §3.3 'Current agricultural potential' (p.10); Discussion §4.5 (p.13) + Table 2."
- claim: "DIRECT fertiliser-consumption offset ~15% (distinct from the 7-34% food-supply metric, and the smaller number): French P excretions total 28.4 kt P/yr against P-fertiliser consumption of 180-200 kt P/yr. Verbatim: 'For 2015–2020, P fertilizer consumption in France was 180–200 kt P·year−1 , so the 28.4 kt P·year−1 excreted could replace ∼15% of the fertilizer consumption.' Global counterpart (Sutton et al. 2013, cited): 4 MtP excreted vs 14-18 MtP consumed ~= 25%; and it is 'consistently been estimated that recycling all human P excretions could cover roughly 20% of current P fertilizer demand' (Mihelcic et al. 2011, cited). [verified verbatim against the raw]"
  source_location: "Results §3.3 'Current agricultural potential' (p.10); Discussion §4.5 (p.13)."
- claim: "CURRENT French recycling rate ~50% (context): of the 39.1 kt P entering the sanitation system, 19.2 kt P (~50%) is used as crop fertiliser, ~35% (14.2 kt P) reaches surface water or is diffusely lost, and ~15% (4.2 kt P) is incinerated or landfilled — despite 75% of WWTP sludge being crop-applied and a national WWTP P-removal efficiency of 80%. Verbatim (abstract): 'only 50% of the excreted P is returned to agroecosystems. This is among the highest rate in Western countries.' [verified verbatim against the raw]"
  source_location: "Abstract; Results §3.1 'French sanitation system P budget', Figure 1 (pp.8-9); Discussion §4.3."
- claim: "METHOD (why 7% vs 34%): the metric is a 'P footprint' method, Pf = Ps/PUE, comparing the P content of excretions to the footprint of domestic food supply, split between plant- and animal-based products because they have very different Phosphorus Use Efficiencies (crops PUE ~50-105% vs animal products ~5-10%; total French food PUE ~11%). Directing recovered P to high-PUE crops for human consumption is what raises the achievable share from 7% to 34% (globally 10-16% -> up to 40%). The framing is food security / reducing national dependence on foreign phosphate rock, not a claim that excreta meets 34% of total fertiliser demand. [verified verbatim against the raw]"
  source_location: "Methods §2.3 'Potential for domestic food consumption'; Discussion §4.5 + Table 2 (p.13); Appendix A.4."

Neobiome Intelligence relevance

Feeds D03 and D02 as a peer-reviewed VALIDATION of the metric framing — that recovered excreta nutrients offsetting a measurable share of fertiliser/food-nutrient need is a recognised, quantified approach — not as a new numeric model input. It is the precedent behind the composting-toilet / digestate nutrient-loop lever (cf. LIT_060, and the Gullberg nutrient-recovery figures on the D03 page).

Two things the model should take from it:

1. Tighten the wording to P-self-sufficiency. The model’s excreta-offset metric should be expressed as a phosphorus food-self-sufficiency share (P in excreta ÷ the P footprint of domestic food supply), not a generic “excreta offsets X% of fertiliser demand”. This source gives no N figure, so any N-side excreta-offset claim needs its own source.
2. Do not conflate the two figures. The evocative 7-34% is the food-supply metric; the direct fertiliser-consumption offset is only ~15% (28.4 vs 180-200 kt P/yr). If the model reports a headline excreta-offset percentage, it must state which of these two it is.

⚠ Transferable proxy, no NZ value. France (metropolitan), not NZ — same status as LIT_060. The percentages are diet- and food-system-specific to France (animal-heavy Western diet); a NZ figure would need NZ diet + food-P-footprint data. Use it for the metric’s shape and defensibility, not its numbers.

Notes

• Accepted manuscript, cite 2024. Filed PDF is the open-access accepted manuscript (HAL / arXiv:2310.06461v2, 4 Jul 2024); the published version is Nutrient Cycling in Agroecosystems (2024) 129:411-425, DOI 10.1007/s10705-024-10367-4. The “2023” in the original download filename is the 2023 preprint — the raw is renamed to the canonical 2024 page stem on filing.
• Phosphorus only. The paper is a P mass balance and gives no nitrogen figure — do not read any N number off it.
• Scope: metropolitan France (incl. Corsica, excl. overseas territories). data_quality high (peer-reviewed; every quoted figure read directly against the raw; the 7-34% and ~15% figures independently corroborated within the paper by the global ~20-25% literature it cites — Mihelcic et al. 2011, Sutton et al. 2013).

Research targets

Documents to retrieve

• None opened.

Research gaps

• NZ counterpart (recorded, not opened): a NZ-specific excreta-nutrient food-self-sufficiency figure (NZ diet + food-P/N footprint) would be the NZ counterpart to this France proxy. Low priority unless the excreta-offset metric becomes a live model output.

Connections

Links to

EDT domains (2): D02: Smart Food Systems & Agriculture · D03: Water, Waste & Circular Systems

Sources (1): LIT_060

Referenced by

EDT domains (2): D02: Smart Food Systems & Agriculture · D03: Water, Waste & Circular Systems