The Expected case projects an annual saving of ₹4.21 crore against a manual baseline cost of ₹302 crore/year — a 1.4% improvement on total material spend. While 1.4% sounds modest in isolation, it represents a 6.7× return on the projected implementation cost over a three-year horizon, and the saving compounds: every batch run after implementation captures the gain.
The confidence band is bracketed at ±46% — placing the portfolio in the Realistic zone of the discovery spectrum. This is the typical position for an early-stage assessment and indicates the Expected number is usable as a working figure for board discussion, with ±₹1.6 crore of movement likely as deeper plant data refines the bracket.
The three highest-contribution blends — HD diesel 15W-40 (CI-4), PCMO 5W-30 (SN/SP), and Hydraulic ISO 68 — are all high-volume products with above-average additive treat rates. The 0.6% gap between the plant's current additive dosing accuracy and OmniBlend's best-practice target is the single largest value driver in the portfolio.
The confidence band measures the width of the saved Low-to-High scenarios relative to Expected. Tighter bands signal higher confidence in the headline. Bharat Lube sits in the Realistic zone — typical for early-stage discovery.
The top-three concentration is the single most important fact for implementation prioritisation. Capture these three blends first and the rest is incremental refinement.
The 67% concentration in three blends is a strategic gift. Rather than spreading optimisation effort across all seven products at once, the plant can run a tightly-scoped pilot on the top three and capture the bulk of the value. This compresses time-to-value from 18 months to roughly 6 months.
The remaining four blends (transformer IS 335, gear ISO 220, hydraulic ISO 46, metalworking semi-synthetic) account for ₹1.38 crore combined and should be implemented in a second wave once the pilot data is in.
Each loss category carries a plant-grade name from the standard mid-segment cost report — additive giveaway, spec giveaway, evaporation, FTR/rework, off-spec/downgrade, cycle-time. Their relative size tells the implementation team which lever moves first.
Additive giveaway accounts for 54% of the total pool — more than the next two categories combined. This is consistent with what mid-segment Group II/III blending plants typically see when they move from manual gravimetric dosing to a real-time predictive density loop.
The single highest-leverage technical change is to tighten additive dosing accuracy on the top three blends from the plant's current ±2.8% manual baseline toward OmniBlend's ±0.6% best-practice target. Cycle-time giveaway (₹0.86 Cr) is the second lever and tracks closely with LIMS-to-blend-tank handoff latency — which is a process design issue rather than an instrumentation issue.
Evaporation and off-spec losses are below ₹0.15 Cr each and should not be prioritised — they are well-controlled at this plant.
Each cell shows the projected annual saving for one blend in one confidence scenario. Darker cells signal larger absolute savings. The vertical pattern of dark cells reveals which blends carry the heavy lifting; the horizontal pattern reveals how much the High scenario adds beyond Expected.
The vertical pattern is striking: the top three rows are several shades darker than the bottom four across every scenario column. This visual concentration confirms the rank-ordering on page 4 — implementation effort focused on HD diesel 15W-40, PCMO 5W-30, and Hydraulic ISO 68 captures the visible savings density.
The horizontal pattern is also informative. Each row's High column is roughly 1.5× the Expected column, indicating consistent upside across the portfolio. This is a healthy signal — it means the High scenarios were bracketed using realistic gap-closure assumptions rather than aspirational ones.
Each recommendation names the action, the expected impact, and the right owner. Implementation should run sequentially — Recommendation 1 unlocks the data needed for Recommendation 2.
None of these risks is a deal-breaker — but each one carries a specific mitigation. The implementation partner should review these with the plant ops head before kickoff.
A ±46% band reads as Realistic. This is the typical position for a 90-minute discovery workshop where scenarios are bracketed against published plant benchmarks rather than measured plant SPC data.
The Expected case is usable as the working figure for board approval, but the spread implies ±₹1.6 crore of legitimate movement as discovery deepens. To tighten the band toward the Tight zone (±20% or less), the plant ops team should validate the Low and High assumptions for the three highest-contribution blends against current LIMS density and COA additive treat-rate data.