Back to the story

Bhimavaram shrimp — supply & production flow

SANKEY · SOURCE: SANKEY.XLSX (22 FLOWS) · HOVER A BAND TO TRACE IT

Product → market split

Show

Product type

Reading this / what the sheet does not say

  1. The product-to-market split in the sheet is uniform, and cannot be right. Every one of the four product types is entered as sending exactly 7.65 bn domestic and 11.475 bn international. So Value-added receives 0.77 bn from the processor but emits 19.125 bn — a 25× gain across a node that only packs. The column totals still add to 76.5 bn, so the error hides in the totals. Switch to Proportional 40/60 to apply the overall domestic:international ratio (30.6 bn : 45.9 bn) to each product instead; every column total stays identical, only the split changes.
  2. Flow is not conserved at two points, and a Sankey assumes it is. NBC 13,222 → BMC 210,000 → hatchery 90 bn is biological multiplication, not transport — a few thousand broodstock become billions of post-larvae. That is drawn here on its own thickness scale, marked by the dashed break, because on the downstream scale 13,222 is 1/7,000,000th of a pixel. Second, the farm takes in 90 bn and passes on 76.5 bn; the 13.5 bn gap is not in the sheet, so it is drawn as a derived unaccounted stub rather than quietly dropped.
  3. Units are not stated in the sheet. Upstream the values are plainly counts of animals (broodstock, then post-larvae). Downstream, 76.5 bn is too large to be kilograms of shrimp and too small to be rupees of export value, so the downstream figures are carried through as entered and labelled only as units. Worth resolving before this goes in the report.
  4. Everything below the processor does balance: the four product types sum to 76.51 bn against 76.5 bn in, and the seven export destinations sum to 45.9 bn against 45.9 bn leaving International.