Post-Harvest Loss Mapping in the Kolar Tomato Supply Chain and the Viability of an FPO Pack-House with Pre-Cooling

Kolar district is one of India's largest tomato-producing belts, and the Kolar APMC is among the biggest tomato markets in the country, sending produce to Bengaluru, Chennai, Hyderabad and beyond. Prices swing from a few rupees a kilo in a glut to headline-making highs in a shortage, and a visible share of the crop is lost between the field and the consumer: crushed in crates, over-ripened while waiting for auction, rejected by retailers.

This dissertation is carried out with a fictional farmer producer company, Kolar Tomato Growers Producer Company Ltd. (about 900 member farmers), which is considering investing in a pack-house with grading, pre-cooling and a small cold room at its aggregation point. The FPO's board wants two answers: where in the chain is tomato actually lost, and how much, and would the investment pay for itself.

The study maps losses at five nodes (harvest and on-farm grading, farm-to-APMC transport, APMC handling and auction, wholesale-to-retail transport, retail) using interviews with 120 farmers, 15 commission agents and traders and 20 Bengaluru retailers, supported by weighed sample checks at each node. It analyses determinants of farm-level loss in SPSS and models the pack-house investment in Python with NPV, IRR, payback and a Monte Carlo simulation of price and utilisation uncertainty, including applicable central schemes for post-harvest infrastructure. The report follows VTU MBA 22MBAPR407 Project Work.

Abstract

This study estimates post-harvest losses of tomato at each node of the Kolar-to-Bengaluru supply chain and evaluates the financial viability of a farmer producer company's pack-house with pre-cooling. Primary data from 120 farmers, 15 traders and 20 retailers, with weighed sample checks, is used to quantify node-wise losses and identify their determinants. A discounted cash-flow model with Monte Carlo simulation tests the investment under price and utilisation uncertainty. The study recommends where the FPO should intervene first and under what conditions the cold-chain investment is viable.

Introduction

Fruits and vegetables suffer higher post-harvest losses than cereals because they are perishable, bulky and handled many times. Studies commissioned by the Government of India, including the ICAR-CIPHET assessment (2015) and the NABCONS study for the Ministry of Food Processing Industries (2022), have estimated national loss levels by crop, but district supply chains differ widely in practices, distances and market structure. Farmer producer organisations are now encouraged to build aggregation and post-harvest infrastructure through schemes such as the Agriculture Infrastructure Fund and the Mission for Integrated Development of Horticulture.

Literature gap

National studies give averages, not the node-level detail an FPO needs. Supply-chain projects by MBA students commonly describe "SCM of Company X" without quantifying losses or testing an investment. Few studies link measured losses at the district level to a financial model of a specific intervention.

Proposed study

• Map the chain and measure loss at each node.
• Identify factors that explain farm-level loss.
• Compute price spread and marketing efficiency.
• Build an investment model for a pack-house with pre-cooling, with sensitivity and simulation.

Feasibility

• Operational: the FPO introduces the student to member farmers and to commission agents at the APMC; fieldwork fits the 6-week project window after Semester 3, timed with the harvest season.
• Technical: SPSS and Python are taught in the programme.
• Ethical: consent from all respondents; no Aadhaar, land-record or bank details collected.

Tomato growers in Kolar lose part of their harvest at several points before it reaches consumers, but neither farmers nor their FPO know how losses are distributed across harvest, grading, transport, auction and retail, or which practices drive them. Without this, the FPO cannot decide whether to invest in crates, grading, transport, pre-cooling or cold storage, and lenders and scheme officials ask for evidence of viability. Prices are volatile, so any investment faces uncertain revenue. The problem for this study is to quantify node-wise post-harvest losses in the Kolar tomato supply chain, identify their determinants, and assess whether an FPO pack-house with pre-cooling is financially viable under realistic price and utilisation scenarios.

The study covers tomato grown by members and non-members around the FPO's aggregation point in Kolar district and sold through Kolar APMC to Bengaluru wholesale and retail outlets during one harvest season. Processing (puree, ketchup) and export channels are excluded. Loss estimates combine weighed samples and recall, and apply to the season studied. The financial model is a pre-feasibility analysis, not a bankable detailed project report.

Research design

Descriptive and analytical, mixed methods: structured interviews and weighed sample checks for loss estimation, plus a financial model of an intervention.

Sampling

Loss measurement

• Node 1 (harvest and grading): farmer recall for the last three pickings plus weighed checks of rejects in 30 crates at 10 farms.
• Node 2 (farm to APMC): crates weighed and inspected on arrival for 40 consignments; damaged fruit sorted and weighed.
• Node 3 (APMC handling and auction): agent recall plus observation of waiting time and dumping on glut days.
• Node 4 (wholesale to retail): retailer recall and inspection of 20 consignments on arrival.
• Node 5 (retail): daily unsold and discarded quantity recorded by retailers for one week. Loss % per node = lost quantity ÷ quantity entering that node; cumulative loss computed multiplicatively; value of loss at node-specific prices.

Analysis

• Descriptives and node-wise loss table; Kruskal-Wallis for loss differences by farm size.
• Multiple regression of farm-level loss % on crate use (vs gunny bags), harvest stage, distance to APMC, waiting time, vehicle type.
• Price spread and Acharya–Agarwal marketing-efficiency index from Agmarknet and retailer prices.

Investment model (Python)

• Capital cost: grading line, pre-cooler, 10–15 tonne cold room, crates, building; operating cost: power, labour, maintenance.
• Benefits: avoided losses (from measured Node 1–3 losses), grade premium, ability to hold produce for 2–5 days in glut periods.
• Financing: own funds plus loan with interest subvention under the Agriculture Infrastructure Fund and any capital subsidy available under MIDH or state schemes (scheme rules checked at the time of study).
• Outputs: NPV at a stated discount rate, IRR, discounted payback; one-way sensitivity on price spread, utilisation and power tariff; Monte Carlo (10,000 runs) drawing price and utilisation from distributions fitted to Agmarknet history.

Timeline (20 weeks)