pith. sign in

arxiv: 2606.11099 · v1 · pith:TUYLC2PLnew · submitted 2026-06-09 · 🧮 math.OC · physics.chem-ph

The Fuel Optimizer: A Data-Driven Numerical Framework for Formulation of Aviation Turbine Fuel

Ana Larra\~naga , Steven L. Brunton , Jacobo Porteiro , Dario Lopez-Pintor This is my paper

Pith reviewed 2026-06-27 12:13 UTC · model grok-4.3

classification 🧮 math.OC physics.chem-ph
keywords aviation turbine fuelfuel formulationsurrogate modelinggenetic algorithmemissions optimizationsustainable aviation fuelsinverse design
0
0 comments X

The pith

A numerical framework uses surrogate models and genetic algorithms to design aviation fuel blends that reduce emissions while meeting property standards.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces the Fuel Optimizer as an inverse design framework for sustainable aviation fuels. It takes a user-defined merit function and finds optimal combinations of chemical species or hydrocarbon families to optimize targets. The method builds a database of qualifying fuel blends, runs reactor simulations to obtain emissions data at cruise conditions, trains a surrogate model that predicts emissions from composition, and applies a genetic algorithm to search for better formulations. Two merit functions target breaking the NOx-CO trade-off and cutting landing-and-take-off cycle emissions, with constraints on property standards including seal swelling and composition limits. The resulting candidates outperformed the original database on the merit functions and were checked with further reactor simulations.

Core claim

The Fuel Optimizer is an inverse design framework that starts from a user-defined merit function and identifies the optimal combination of chemical species or hydrocarbon families that optimize a combination of targets. As a case study, a database of fuel blends meeting selected property standards was simulated in a reactor model to obtain pollutant emissions at cruise conditions. A surrogate model was developed to reduce the computational cost of evaluating candidate blends, taking fuel composition as input and predicting emissions as output. A genetic algorithm was used to optimize fuel formulation according to two merit functions designed to break the nitrogen oxides - CO trade-off, and m

What carries the argument

Surrogate model trained on reactor simulation outputs, paired with genetic algorithm search under merit functions and property constraints.

If this is right

  • Fuel candidates generated by the framework outperformed the training database across all merit functions.
  • Optimal candidates were validated through additional reactor simulations confirming the predicted improvements.
  • The framework supports different user-defined merit functions to address varied emission or performance targets.
  • Property standards including seal swelling and composition limits are enforced as constraints during the search.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The method could speed up exploration of sustainable aviation fuel compositions that lie outside manually curated databases.
  • It might be combined with additional reactor models or experimental validation loops to cover more flight conditions.
  • Analogous inverse-design loops could be applied to optimize other fuel types or chemical mixtures in energy systems.

Load-bearing premise

The surrogate model trained on the initial database accurately predicts emissions for compositions outside that database, and the chosen merit functions plus property constraints correctly represent real operational and regulatory requirements.

What would settle it

Run the reactor simulations on the specific optimal fuel candidates produced by the framework and verify whether their emissions match the surrogate predictions while improving on the training database.

Figures

Figures reproduced from arXiv: 2606.11099 by Ana Larra\~naga, Dario Lopez-Pintor, Jacobo Porteiro, Steven L. Brunton.

Figure 1
Figure 1. Figure 1: This work consists of two separate tasks. The first one corresponds to the offline training of the surrogate model (orange), which includes the generation of a fuel blend database containing all chemical species in the pure components database and the subsequent training of the neural network. The second stage corresponds to the online optimization procedure, in which the trained surrogate model is embedde… view at source ↗
Figure 2
Figure 2. Figure 2: This study explores two fuel blend optimization scenarios: (1) fuel blends that contain oxygenates (purple), and (2) fuel blends without oxygenates but subject to additional constraints (green). The second case is a more practical optimization scenario while the first is more exploratory. Two merit functions, MF1 and MF2, are optimized. MF1 focuses on CO-NOx trade-off and MF2 focused on LTO regulated emiss… view at source ↗
Figure 3
Figure 3. Figure 3: Comparison between the brute-force database (green, labeled DB) and the feasible candidate blends evaluated by the GA (blue, labeled GA) for all optimization scenarios. The left panels correspond to the CO–NOx trade-off (MF1), whereas the right panels correspond to the LTO regulated-emissions objective function (MF2). The bottom panels represent the practical approach, in which oxygenates are excluded and … view at source ↗
Figure 4
Figure 4. Figure 4: Results for the exploratory approach analyzed by chemical family, focusing on the most relevant families included in the database (additional results are provided in the Supplementary Material). Figure (a) on the left corresponds to the CO–NOx trade-off optimization, whereas figure (b) on the right corresponds to the LTO regulated-emissions optimization. The merit-function value is plotted on the y-axis, w… view at source ↗
Figure 5
Figure 5. Figure 5: Results for the practical approach analyzed by chemical family, considering that in this case the database is limited to not include oxygenates. Figure (a) on the top corresponds to the CO–NOx trade-off optimization, whereas figure (b) on the bottom corresponds to the LTO regulated-emissions optimization. The merit-function value is plotted on the y-axis, while the proportion of each chemical family in the… view at source ↗
Figure 6
Figure 6. Figure 6: Detailed analysis of the 50 best unique fuel blends by chemical family (left panels) and by individual chemical species (right panels) for the CO–NOx trade-off objective function. The top panels correspond to the exploratory approach, which considers the complete database, whereas the bottom panels correspond to the practical approach, in which oxygenates are excluded and additional constraints on olefin c… view at source ↗
Figure 7
Figure 7. Figure 7: Detailed analysis of the 50 best unique fuel blends by chemical family (left panels) and by individual chemical species (right panels) for the LTO regulated emissions objective function. The top panels correspond to the exploratory approach, which considers the complete database, whereas the bottom panels correspond to the practical approach, in which oxygenates are excluded and addi￾tional constraints on … view at source ↗
Figure 8
Figure 8. Figure 8: Results for the optimal fuel candidates identified by the Fuel Optimizer for MF1 (CO–NOx trade-off) and MF2 (LTO regulated emissions), evaluated under both exploratory and practical scenarios. The selected blends were simulated using the 1D reactor net￾work under LTO conditions and compared against both experimental and simulated Jet A data, which serves as the baseline reference for performance improvemen… view at source ↗
Figure 1
Figure 1. Figure 1: Comparison between reactor engine model simulations and experimental measurements of a CFM56-7B27 engine at LTO conditions with Jet A. Figure includes combustor outlet temperature (top), NOx and CO emissions (middle), and particulate mass and particle number emissions (bottom) 2 [PITH_FULL_IMAGE:figures/full_fig_p020_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Surrogate model accuracy for the exploratory approach, where the database was not limited. The results are focused on the target variables relevant to the merit functions that are tested in this work. 3 [PITH_FULL_IMAGE:figures/full_fig_p021_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Surrogate model accuracy for the practical approach, where oxygenates were eliminated from the database. The results are focused on the target variables relevant to the merit functions that are tested in this work. Practical Fuel Optimizer. This surrogate follows the same training framework as the exploratory model (Optuna TPE, 100 trials, 5-fold cross-validation, identical callbacks), differing in three a… view at source ↗
Figure 4
Figure 4. Figure 4: Full results for the exploratory approach analyzed by chemical family. 6 [PITH_FULL_IMAGE:figures/full_fig_p024_4.png] view at source ↗
read the original abstract

The Fuel Optimizer is an inverse design framework for sustainable aviation fuels that starts from a user-defined merit function and identifies the optimal combination of chemical species or hydrocarbon families that optimize a combination of targets. As a case study, a database of fuel blends meeting selected property standards was simulated in a reactor model to obtain pollutant emissions at cruise conditions. A surrogate model was developed to reduce the computational cost of evaluating candidate blends, taking fuel composition as input and predicting emissions as output. A genetic algorithm was used to optimize fuel formulation according to two merit functions designed to break the nitrogen oxides - CO trade-off, and minimize pollutant emissions over the landing-and-take-off cycle. Constraints included selected property standards (including seal swelling) and composition limits. Fuel candidates from the framework outperformed the training database across all merit functions, and the optimal candidates were validated through reactor simulations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 0 minor

Summary. The manuscript introduces the Fuel Optimizer, an inverse-design framework for aviation turbine fuels. Starting from a user-specified merit function, it builds a database of property-compliant blends, simulates them in a reactor model to obtain cruise emissions, trains a surrogate to map composition to emissions, and applies a genetic algorithm to optimize blends under property and composition constraints. Two merit functions are used to break the NOx-CO trade-off and minimize LTO-cycle emissions. The abstract states that the resulting candidates outperform the training database on all merit functions and that the optima were validated by reactor simulations.

Significance. If the surrogate generalizes reliably outside the training support and the merit functions faithfully capture operational requirements, the framework could provide a practical, data-driven route to low-emission sustainable aviation fuel formulations that would otherwise require prohibitive numbers of reactor evaluations.

major comments (2)
  1. The central claim—that GA-optimized compositions outperform the training database—rests on the surrogate accurately predicting emissions for points outside the initial database. The abstract provides no quantitative surrogate metrics (RMSE, R², cross-validation scores), no out-of-distribution test set, and no uncertainty quantification, so it is impossible to determine whether the reported outperformance is an artifact of surrogate error.
  2. No sensitivity analysis is reported for the chosen merit functions or the property constraints (including seal-swelling). If small changes in the weighting or constraint bounds alter the location of the optima, the superiority claim cannot be considered robust.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments. We provide point-by-point responses to the major comments below, indicating the revisions we will make to address them.

read point-by-point responses

  1. Referee: The central claim—that GA-optimized compositions outperform the training database—rests on the surrogate accurately predicting emissions for points outside the initial database. The abstract provides no quantitative surrogate metrics (RMSE, R², cross-validation scores), no out-of-distribution test set, and no uncertainty quantification, so it is impossible to determine whether the reported outperformance is an artifact of surrogate error.

    Authors: We appreciate this observation. The manuscript does report that the optimal candidates were validated through direct reactor simulations, which provides independent confirmation of their performance and mitigates concerns about surrogate error for the final claims. However, we agree that the abstract should include quantitative surrogate metrics to better support the framework description. In the revised manuscript, we will update the abstract to report key surrogate performance indicators such as RMSE, R², and cross-validation scores. We will also add details on any out-of-distribution testing and uncertainty quantification present in the full text or perform additional analysis if needed. revision: partial

  2. Referee: No sensitivity analysis is reported for the chosen merit functions or the property constraints (including seal-swelling). If small changes in the weighting or constraint bounds alter the location of the optima, the superiority claim cannot be considered robust.

    Authors: We concur that sensitivity analysis would enhance the robustness of our findings. We will incorporate a sensitivity analysis in the revised manuscript, examining the effects of variations in merit function weights and property constraint bounds, with particular attention to seal-swelling properties, to verify the stability of the optimized compositions. revision: yes

Circularity Check

0 steps flagged

No circularity; surrogate optimization followed by independent reactor validation

full rationale

The derivation chain uses a reactor model to generate training data, fits a surrogate, runs GA optimization under merit functions and constraints, then validates the resulting candidates with the original reactor model. This validation step is external to the surrogate fit and to the training database, so the claim that candidates outperform the database is not forced by construction. No self-citations, self-definitional equations, or fitted inputs renamed as predictions appear in the provided text. The process is a standard data-driven optimization pipeline with post-hoc verification.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review yields no explicit free parameters, axioms, or invented entities; all such elements would require the full text.

pith-pipeline@v0.9.1-grok · 5688 in / 983 out tokens · 17802 ms · 2026-06-27T12:13:43.220838+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

12 extracted references · 6 canonical work pages

  1. [1]

    C., Faulhaber, C., Borland, R., Boehm, J., Heyne, Measurements of Nitrile Rubber Absorption of Hydrocarbons: Trends for Sustainable Aviation Fuel Compatibility, Energy and Fuels 37 (2023) 9207–

    2023

  2. [2]

    doi:10.1021/acs.energyfuels.3c00781

    work page doi:10.1021/acs.energyfuels.3c00781

  3. [3]

    Low Lunar Orbit Design via Graphical Manipulation of Eccentricity Vector Evolution

    J., Hamilton, B., Khandelwal, Effects of Aromatic Blends on Seal Swell Rates Using Novel Seals Compatibility Test Rig, AIAA SCITECH 2023 Forum, paper number 0596 (2023). doi:10.2514/6. 2023-0596

    work page doi:10.2514/6 2023

  4. [4]

    URL:https://www.easa.europa.eu/en/ domains/environment/icao-aircraft-engine-emissions-databank

    ICAO Aircraft Engine Emissions Databank | EASA. URL:https://www.easa.europa.eu/en/ domains/environment/icao-aircraft-engine-emissions-databank

  5. [5]

    Spatially Resolving Electron Spin Resonance ofπ-Radical in Single-Molecule Magnet

    Lopez-Pintor, D., MacDonald, J., Ramirez-Correa, E., Garcia-Oliver, J. M., Payri, R., Marti Gomez- Aldaravi, P ., Leyland, P ., Fuel Effects on Aviation Engine Emissions: A Chemical Reactor Network Modeling Study, Energy & Fuels 40 (2026) 4777–4792. URL:https://doi.org/10.1021/acs. energyfuels.5c05727. doi:10.1021/acs.energyfuels.5c05727

    work page doi:10.1021/acs 2026

  6. [6]

    URL:https://store.astm.org/d1655-25a

    Standard Specification for Aviation Turbine Fuels. URL:https://store.astm.org/d1655-25a. html

  7. [7]

    URL: https://wtt-pro.nist.gov/wtt-pro/#default/B;869,179/A;0,0,705,755;help, about/

    NIST/TRC Web Thermo Tables (WTT): Critically Evaluated Thermophysical Property Data. URL: https://wtt-pro.nist.gov/wtt-pro/#default/B;869,179/A;0,0,705,755;help, about/

  8. [8]

    P ., THE VISCOSITY OF LIQUIDS

    Kendall, J., Monroe, K. P ., THE VISCOSITY OF LIQUIDS. II. THE VISCOSITY-COMPOSITION CURVE FOR IDEAL LIQUID MIXTURES.1, Journal of the American Chemical Society 39 (1917) 1787–

    1917

  9. [9]

    doi:10.1021/ja02254a001

    URL:https://doi.org/10.1021/ja02254a001. doi:10.1021/ja02254a001

    work page doi:10.1021/ja02254a001

  10. [10]

    URL:https://doi.org/10

    Lemmon, E.W., Bell, I.H., Huber, M.L., McLinden, M.O., NIST Standard Reference Database 23: Ref- erence Fluid Thermodynamic and Transport Properties-REFPROP , Version 10.0, National Institute of Standards and Technology, Standard Reference Data Program (2018). URL:https://doi.org/10. 18434/T4/1502528. doi:10.18434/T4/1502528

    work page doi:10.18434/t4/1502528 2018

  11. [11]

    URL:https://arc.aiaa.org/ doi/10.2514/6.2022-2221

    Coburn, A., Yang, Z., Boehm, R., Heyne, J., Determination of a Freeze Point Blend Prediction Model for Jet Fuel Range Hydrocarbons, in: AIAA SciTech Forum, 2022. URL:https://arc.aiaa.org/ doi/10.2514/6.2022-2221. doi:10.2514/6.2022-2221

    work page doi:10.2514/6.2022-2221 2022

  12. [12]

    O., Chittenden, D

    Wickey, R. O., Chittenden, D. H., Flash points of blends correlated., Hydrocarb Process Pet Refin 42 (1963) 157–158. 8

    1963