Improving quality of Cofea Arabica L. through addition of sugarcane-based supplements in the anaerobic fermentation process

Date

7-2026

Abstract

This study evaluated the effects of sugarcane-based supplements on the anaerobic fermentation of Coffea Arabica L. and developed a machine learning model to predict coffee cup score. Specifically, the study determined the effects of the type of natural supplement and coffee-to-supplement ratio on the physicochemical characteristics, fermentation behavior, and cup quality of Arabica coffee. A 3x3 factorial experiment arranged in a Completely Randomized Design (CRD) was used. Factor A consisted of fresh sugarcane juice, fermented sugarcane juice and distilled water (control) while Factor B consisted of coffee-to-supplement ratios of 1:½, 1:1, and 1:1½. Data were analyzed using two-way analysis of variance (ANOVA), Tukey’s HSD Test, and Pearson’s correlation analysis. The anaerobic fermenter was designed to facilitate controlled operation and continuous monitoring throughout the fermentation process.

The results showed that sugarcane-based supplements significantly influenced the physiochemical characteristics, fermentation behavior, microbial growth, and cup quality of Arabica coffee. The treatments significantly affected pH, color, crude fiber, ash, crude protein, fermentation time, pH decline rate, carbon dioxide (CO2) production rate, and microbial growth. Fresh and fermented sugarcane supplements promoted greater microbial activity than the control. Coffees fermented with sugarcane supplements obtained higher cup scores and more desirable sensory attributes than those fermented with distilled water. Fresh sugarcane enhanced sweetness and fruity notes, whereas fermented sugarcane produced citrus, fruity and spicy characteristics. In contrast, the control exhibited lower cup scores and greater astringency. The highest cup score was obtained using fresh sugarcane at a coffee-to-supplement ratio of 1:1. Pearson’s correlation analysis showed that the cup score was positively associated with pH and ash content. Among the machine learning models evaluated, Elastic Net achieved the best predictive performance (R2=0.68, RMSE=0.50, and MAE=0.40). Variable importance analysis identified pH decline rate, microbial growth, crude protein, ash content, fermentation time and CO2 production as the most important predictors of cup score.

The study demonstrated that sugarcane-based supplements improved the quality of Cofea Arabica L. during anaerobic fermentation under controlled process operation and monitoring. Furthermore, the developed machine learning model successfully predicted coffee cup quality using fermentation and physiochemical parameters.

Document Type

Dissertation

Degree

Doctor of Philosophy in Agricultural Engineering

Adviser/Committee Chair

Arnold R. Elepaño

Committee Member

Ernesto P. Lozada, Kevin F. Yaptenco, Zenith O. Larnejo

Language

English

En – AGROVOC descriptors

COFFEA ARABICA; SUGAR CANE; FOOD SUPPLEMENTS; ANAEROBIC FERMENTATION; CHEMICOPHYSICAL PROPERTIES; FOOD QUALITY

This document is currently not available here.

Share

COinS