Third infusion falls off a cliff

[Garden In Order]

Question or objective: extend thick texture through at least six infusions
Tea category: ripe puer
Exact product: Old Brick Courtyard 2022 shou
Vendor: Cloudstep Tea
Origin: Menghai, Yunnan
Cultivar, if known: large-leaf assamica
Harvest season/year: 2021 material / pressed 2022
Processing or roast: wo dui fermented, compressed cake
Cake/batch number, if relevant: cake/lot noted above where applicable; otherwise N/A
Purchase/opening date: purchased 2026-06-07; opened 2026-07-04
Storage since purchase: breathable paper inside a stable tea cabinet unless otherwise noted
Style: gongfu
Vessel: gaiwan
Vessel material: porcelain
Nominal and actual filled capacity: 110 ml nominal / 100 ml filled
Filter or strainer: gaiwan lid
Kettle: standard electric kettle + thermometer
Vessel preheated: yes
Dry leaf weight: 7.0 g
Water volume: 100 ml
Water temperature: 99°C
Water source or mineral profile: filtered tap water, Portland; about 70 ppm
Rinse: 5-10 s where noted by the recipe; otherwise none
Infusion schedule: 10 s rinse, 12 s, 15 s, 20 s, 30 s
Agitation: gentle shake after rinse only
Was each infusion completely decanted?: yes
Aroma: wood, cocoa
Sweetness: medium
Bitterness: none
Astringency: low
Body/texture: full then suddenly thin
Flavour: cocoa, old wood, date
Aftertaste: short after infusion three
Number of satisfying infusions: 4
Exact problem: tea loses body abruptly after the third infusion
Desired result: extend thick texture through at least six infusions
Variable changed: leaf quantity
Previous value: 6 g
New value: 7 g
Result: first two steeps got heavier but the drop-off still happened
Dry-leaf photograph: available in the author’s notes; not embedded in this archive
Liquor photograph: available in the author’s notes; not embedded in this archive
Wet-leaf photograph: available in the author’s notes; not embedded in this archive
Video if pouring or drainage is relevant: pour/drainage video described in the post; not embedded in this archive

I have reached the stage where changing three things at once is only producing new kinds of confusion. The cup is giving me tea loses body abruptly after the third infusion. The last controlled change was leaf quantity: 6 g to 7 g, and first two steeps got heavier but the drop-off still happened. I would like the next test to be one variable only. If you have brewed this exact tea, or something genuinely comparable in a similar vessel, that would be especially useful.

My target is simple: extend thick texture through at least six infusions. I am not looking for a new shopping list unless the equipment is actually the problem.

[Garden In Order]

Tested solution
Direct recommendation: add a 20-second rest after the rinse before the first infusion.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: firsthand test by the questioner on Old Brick Courtyard 2022 shou.
Change: add a 20-second rest after the rinse before the first infusion.
Expected effect: more even extraction and a slower falloff.
Observed result: first two steeps got heavier but the drop-off still happened.
Limitations: My result is from a small number of sessions, so treat it as evidence rather than law.
Commercial relationship: none.

[Quiet Stacks]

Comparable tea
Direct recommendation: extend infusion three rather than increasing leaf.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: firsthand use of a comparable ripe puer with similar parameters.
Change: extend infusion three rather than increasing leaf.
Expected effect: better body in the middle infusions without an over-heavy first cup.
Observed result: not tested on this exact setup yet.
Limitations: Water chemistry may change the result, so I would not generalize this to every city.
Commercial relationship: none.

[Masala Soil]

Professional experience
Direct recommendation: extend infusion three rather than increasing leaf.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: general brewing principle consistent with repeated tests; not this exact batch.
Change: extend infusion three rather than increasing leaf.
Expected effect: better body in the middle infusions without an over-heavy first cup.
Observed result: not tested on this exact setup yet.
Limitations: My result is from a small number of sessions, so treat it as evidence rather than law.
Commercial relationship: none.

[Masala Soil]

Same tea/batch
Direct recommendation: break the chunk apart more before brewing.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: firsthand use of the same Old Brick Courtyard 2022 shou in a closely matched setup.
Change: break the chunk apart more before brewing.
Expected effect: more consistent body across infusions.
Observed result: not tested on this exact setup yet.
Limitations: My result is from a small number of sessions, so treat it as evidence rather than law.
Commercial relationship: none.

[Quiet Stacks]

Comparable tea
Direct recommendation: break the chunk apart more before brewing.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: firsthand experience with ripe puer in a different vessel.
Change: break the chunk apart more before brewing.
Expected effect: more consistent body across infusions.
Observed result: not tested on this exact setup yet.
Limitations: This may not transfer to a heavier roast or a very different vessel size.
Commercial relationship: none.

[Masala Soil]

Controlled test
Direct recommendation: add a 20-second rest after the rinse before the first infusion.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: side-by-side or repeated one-variable test.
Change: add a 20-second rest after the rinse before the first infusion.
Expected effect: more even extraction and a slower falloff.
Observed result: not tested on this exact setup yet.
Limitations: My result is from a small number of sessions, so treat it as evidence rather than law.
Commercial relationship: none.

[Garden In Order]

Professional experience
Direct recommendation: extend infusion three rather than increasing leaf.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: general brewing principle consistent with repeated tests; not this exact batch.
Change: extend infusion three rather than increasing leaf.
Expected effect: better body in the middle infusions without an over-heavy first cup.
Observed result: not tested on this exact setup yet.
Limitations: My result is from a small number of sessions, so treat it as evidence rather than law.
Commercial relationship: none.

[Quiet Stacks]

Inference
Direct recommendation: add a 20-second rest after the rinse before the first infusion.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: inference from the posted parameters, not a test of this exact tea.
Change: add a 20-second rest after the rinse before the first infusion.
Expected effect: more even extraction and a slower falloff.
Observed result: not tested on this exact setup yet.
Limitations: This may not transfer to a heavier roast or a very different vessel size.
Commercial relationship: none.

[Garden In Order]

Controlled test
Direct recommendation: break the chunk apart more before brewing.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: side-by-side or repeated one-variable test.
Change: break the chunk apart more before brewing.
Expected effect: more consistent body across infusions.
Observed result: not tested on this exact setup yet.
Limitations: This may not transfer to a heavier roast or a very different vessel size.
Commercial relationship: none.

[Garden In Order]

Controlled test
Direct recommendation: add a 20-second rest after the rinse before the first infusion.
Applies to: Old Brick Courtyard 2022 shou / ripe puer in a setup close to gaiwan.
Evidence: side-by-side or repeated one-variable test.
Change: add a 20-second rest after the rinse before the first infusion.
Expected effect: more even extraction and a slower falloff.
Observed result: not tested on this exact setup yet.
Limitations: This may not transfer to a heavier roast or a very different vessel size.
Commercial relationship: none.

[Author]

Posts