BD × AI Lab
DGTE / HTH / MGHB · Arena v0.7a · Bench planner

DGTE Arena Bench Planner

A public evidence page translating v0.6b/v0.7a arena signals into the smallest physical experiments that can kill, salvage, or advance DGTE, HTH, MGHB, and hybrid branches.

Status: this is not a product-readiness claim. v0.7a generated 42 bench plans from the latest v0.6b baseline (baseline_zip:DGTE_v06_MICRO_results_20260517_164121.zip; rows:2100). Product gate remains hard-failed by design. The next move is measurement.
42
bench plans
7
device families
6
water-bell helpers
0
product claims
Current arena verdict

What v0.7a says right now

The bench planner should be read as a ranked falsification map, not as a proof that any device is market-ready.

Build first

HTH / trompe / pressure

Strongest low-cost kill-test branch. First answer the clear-pipe entrainment and separator-pressure questions.

High upside

MGHB / wave / breakwater

Highest infrastructure upside, but it needs a wave tank to prove wave reduction plus useful work.

Alive but hard

Pure DGTE core

Do not build the full loop first. Attack rotary/starwheel crossover, hydraulic lock, and thermal slice separately.

Helper only

Siphon / thermosiphon / bell

Useful for priming, circulation, anti-air intake, or wet-lock assist. Never source credit.

Strict comparator

Solar DGTE

Likely loses to PV/PV‑T unless it upgrades special waste heat or a site-specific thermal store.

Caution

Hybrid value

Modular tandem is safer than over-integrated claims until parent-only baselines are beaten after penalties.

Build order

Experiment 0 + first 5 physical tests

Experiment 0 is the cheap water-sealed bell side-bench. The first five are staged parent/kill-test builds before integrated hybrid claims.

0raw rank 17 · score 67.9

Water-sealed bell / siphon-primer bench

Experiment 0 / flooded-cup primer side-bench · Siphon helper · VERY_LOW (€50–€250) · ~1 days

Question: Does an inverted water-filled cup/bell over a hose inlet reduce priming loss, prevent air ingestion, and improve shallow-water pickup versus a bare hose?

Measure

  • start success rate
  • time to prime
  • flow rate Q liters/min
  • outlet head Δh
  • minimum water depth before prime loss
  • air ingestion rate / visible bubbles
  • pressure drop / loss coefficient
  • remaining water depth after drain
  • debris clogging tendency

Kill criteria

  • loses prime easily
  • flow rate too low
  • adds more hydraulic loss than it saves
  • clogs or fouls too easily
  • requires too much manual setup
  • does not reduce pump-start energy or transfer J/cycle
  • only works in shallow demo conditions but fails under repeated cycling

Accounting: HELPER_ONLY: WATER_SEALED_BELL_INTAKE is a passive hydraulic helper; source_credit_kw=0; credit only saved loss/J after added loss and complexity. source_credit_kw=0: this primitive is never an energy source. | Credit only reduced priming loss, reduced pump-start energy, reduced air-ingestion loss, or reduced transfer J/cycle. | A candidate only benefits when net_helper_value = saved_loss_or_saved_J - added_loss - added_complexity_penalty is positive. | Do not let this primitive improve product gate by itself; repeated-cycle reliability must pass first.

1raw rank 1 · score 85.1

Transparent HTH / trompe channel

minimum clear-pipe velocity ramp · HTH / trompe · VERY_LOW (€50–€250) · ~2 days

Question: What velocity/head first creates stable entrainment and positive separator pressure?

Measure

  • flow velocity
  • entrainment threshold
  • bubble survival distance
  • separator pressure
  • air flow rate
  • water head / depth
  • net pneumatic power
  • hydraulic/parasitic losses

Kill criteria

  • no stable entrainment below practical velocities
  • separator pressure too low for useful air output
  • parasitic losses exceed useful pneumatic output
  • bubble transport collapses before separator

Accounting: EXTERNAL_RESOURCE: river/current/depth head measured; no gravity/buoyancy free-source credit. External energy is river/current/depth head, not gravity as free source. | Report pneumatic output after water-flow and separator losses. | Do not credit downstream water flow twice if an air turbine is measured.

2raw rank 6 · score 74.5

MGHB wave tank cell

piston-only baseline · MGHB / wave · LOW (€250–€900) · ~5 days

Question: What fraction of incoming wave energy can a guided piston convert without pneumatic hybrid complexity?

Measure

  • incoming wave height
  • outgoing wave height
  • piston stroke
  • piston force
  • generator/load output
  • damping setting
  • water leakage/slam
  • wave period
  • absorbed wave fraction

Kill criteria

  • wave reduction poor
  • useful work fraction too low
  • mechanical loads too high
  • hybrid complexity worse than piston-only or pneumatic-only
  • storm/surge mode cannot fail safe

Accounting: EXTERNAL_RESOURCE: incoming wave energy measured and normalized against wave reduction/output. External source is wave energy crossing the cell, not buoyancy alone. | Credit either piston output or OWC pneumatic output separately before hybrid sum. | Compare against passive breakwater wave reduction and piston-only baseline.

3raw rank 5 · score 77.3

Pure DGTE rotary/starwheel crossover rig

single pocket dye/carry-over test · Pure DGTE · VERY_LOW (€50–€250) · ~3 days

Question: Can a wetted starwheel transfer capsules with low carry-over?

Measure

  • transfer J/cycle
  • carry-over fluid mass
  • leakage
  • torque/current per transfer
  • pressure balance stability
  • cycle rate
  • jam rate
  • seal friction

Kill criteria

  • transfer work exceeds expected stroke gain
  • carry-over too high
  • sealing friction too high
  • cycle rate too low
  • pressure-balanced pockets cannot stay repeatable

Accounting: THERMAL_CLOSURE: P_net=min(P_thermal_gross,P_PTO_path)-transfer-auxiliary. Crossover does not create energy; it must reduce losses enough for thermal stroke to matter. | Measured J/transfer must be compared against expected per-cycle stroke work. | Carry-over heat/fluid penalty must be included.

4raw rank 11 · score 69.9

Hydraulic balanced lock rig

empty lock equalization baseline · Pure DGTE · VERY_LOW (€50–€250) · ~3 days

Question: What is the minimum pressure-equalization cost before capsules are added?

Measure

  • pressure equalization energy
  • lock leakage
  • valve losses
  • throughput
  • J/transfer
  • fouling sensitivity
  • timing jitter

Kill criteria

  • lock overhead kills net power
  • fouling or complexity too high
  • valve losses dominate
  • throughput too low for useful cycle rate

Accounting: THERMAL_CLOSURE: P_net=min(P_thermal_gross,P_PTO_path)-transfer-auxiliary. Pressure equalization is not free; count valve and timing losses. | Lock only survives if it reduces transfer loss below stroke-gain budget. | No product claim until fouling and repeatability are tested.

5raw rank 27 · score 63.3

Integrated DGTE thermal slice

controlled heater closure · Pure DGTE · MEDIUM (€900–€2800) · ~6 days

Question: Can a DGTE slice close thermal-to-mechanical accounting with a known heat input?

Measure

  • hot/cold temperatures
  • thermal input
  • real ΔT
  • cycle work
  • transfer losses
  • auxiliary pumping
  • net electrical output
  • heat exchanger area
  • recuperator effectiveness

Kill criteria

  • thermal-to-mechanical closure fails
  • required heat exchanger area too large
  • low-grade waste heat insufficient
  • auxiliary pumping exceeds generated work
  • cycle rate too low

Accounting: THERMAL_CLOSURE: P_net=min(P_thermal_gross,P_PTO_path)-transfer-auxiliary. Use P_net = min(P_thermal_gross, P_mechanical_or_PTO_path_gross) - transfer losses - auxiliaries. | Waste heat must pay exergy/temperature limit. | Heat source and sink must be measured as a pair.

Bench priority

Top 20 ranked plans

RankExperimentFamilyScorePriorityCostLearn/€
1Transparent HTH / trompe channel
minimum clear-pipe velocity ramp
HTH / trompe85.1A_BUILD_FIRSTVERY_LOW
€50–€250
1.000
2Transparent HTH / trompe channel
separator pressure + loaded airflow
HTH / trompe81.4A_BUILD_FIRSTLOW
€250–€900
0.891
3Transparent HTH / trompe channel
deep-pool/polyline emulator
HTH / trompe80.6A_BUILD_FIRSTLOW
€250–€900
0.886
4Transparent HTH / trompe channel
debris/fouling ruggedness pass
HTH / trompe80.2A_BUILD_FIRSTLOW
€250–€900
0.872
5Pure DGTE rotary/starwheel crossover rig
single pocket dye/carry-over test
Pure DGTE77.3A_BUILD_FIRSTVERY_LOW
€50–€250
0.956
6MGHB wave tank cell
piston-only baseline
MGHB / wave74.5B_STRONG_BENCHLOW
€250–€900
0.792
7Pure DGTE rotary/starwheel crossover rig
cycle-rate limit test
Pure DGTE72.2B_STRONG_BENCHLOW
€250–€900
0.795
8Pure DGTE rotary/starwheel crossover rig
torque/current sweep
Pure DGTE72.1B_STRONG_BENCHLOW
€250–€900
0.795
9MGHB wave tank cell
OWC/pneumatic-only baseline
MGHB / wave70.4B_STRONG_BENCHMEDIUM
€900–€2800
0.687
10Water-sealed bell / siphon-primer bench
DGTE wet-crossover bell helper
Pure DGTE70.0B_STRONG_BENCHVERY_LOW
€50–€250
0.963
11Hydraulic balanced lock rig
empty lock equalization baseline
Pure DGTE69.9B_STRONG_BENCHVERY_LOW
€50–€250
0.878
12Water-sealed bell / siphon-primer bench
HTH flooded intake stabilizer
HTH / trompe69.8B_STRONG_BENCHVERY_LOW
€50–€250
0.967
13Water-sealed bell / siphon-primer bench
MGHB water-filled coupling / anti-air seal
MGHB / wave69.4B_STRONG_BENCHVERY_LOW
€50–€250
0.963
14Water-sealed bell / siphon-primer bench
DGTE-HTH passive wet-lock assist line
DGTE × HTH69.3B_STRONG_BENCHVERY_LOW
€50–€250
0.958
15MGHB wave tank cell
slam/leak survival cycle
MGHB / wave69.3B_STRONG_BENCHMEDIUM
€900–€2800
0.681
16DGTE-HTH modular assist slice
modular tandem decision pass
DGTE × HTH68.5B_STRONG_BENCHMEDIUM
€900–€2800
0.665
17Water-sealed bell / siphon-primer bench
Experiment 0 / flooded-cup primer side-bench
Siphon helper67.9B_STRONG_BENCHVERY_LOW
€50–€250
0.979
18DGTE-HTH modular assist slice
pressure buffer only
DGTE × HTH67.8B_STRONG_BENCHMEDIUM
€900–€2800
0.665
19Pure DGTE rotary/starwheel crossover rig
double-rotary cassette comparison
Pure DGTE67.4B_STRONG_BENCHMEDIUM
€900–€2800
0.676
20DGTE-HTH modular assist slice
degassing + pressure smoother
DGTE × HTH67.2B_STRONG_BENCHMEDIUM
€900–€2800
0.654
Device families

Family-level status

FamilyPlansBest current experimentBest scoreMean riskCost classesVerdict
HTH / trompe
Build first
5Transparent HTH / trompe channel
minimum clear-pipe velocity ramp
85.10.294VERY_LOW:2, LOW:3Strongest low-cost kill-test branch: transparent channel, entrainment threshold, separator pressure, loaded airflow, and losses.
MGHB / wave
High upside
6MGHB wave tank cell
piston-only baseline
74.50.388LOW:1, MEDIUM:2, VERY_LOW:2, HIGH:1Best coastal/infrastructure upside, but needs wave-tank proof of wave reduction plus useful work.
Pure DGTE
Alive but hard
13Pure DGTE rotary/starwheel crossover rig
single pocket dye/carry-over test
77.30.462VERY_LOW:3, LOW:5, MEDIUM:4, HIGH:1Do not jump to a full machine. Attack crossover work, carry-over, leakage, cycle rate, and thermal closure separately.
DGTE × HTH
Modular only
5Water-sealed bell / siphon-primer bench
DGTE-HTH passive wet-lock assist line
69.30.448VERY_LOW:1, MEDIUM:3, HIGH:1Keep as modular tandem. Hybrid value must beat best parent modes after input and complexity penalty.
Siphon helper
Helper only
5Water-sealed bell / siphon-primer bench
Experiment 0 / flooded-cup primer side-bench
67.90.252VERY_LOW:5Cheap circulation/priming branch. It can reduce losses but cannot be counted as an energy source.
Solar/PV compare
Comparator
4Solar/PV-water comparison loop
PV counterfactual same-area test
54.90.360LOW:1, MEDIUM:3Solar DGTE survives only if it beats same-area PV/PV‑T or uniquely upgrades low-grade waste heat.
Frontier
Seed bank
4Frontier membrane/bubble-pod rig
single semi-sealed pod
37.10.795VERY_LOW:1, LOW:3Keep for isolated curiosity tests and salvage; do not let it outrank cleaner kill-tests without data.
v0.7a helper primitive

Water-sealed bell / flooded-cup primer

Passive helper

What it is

The flooded-cup trick becomes WATER_SEALED_BELL_INTAKE: an inverted water-filled cup/bell over a tube inlet that helps prime the tube, keeps air out, allows shallow-water pickup, and stabilizes intake flow.

Best v0.7a rank: #10 — Water-sealed bell / siphon-primer bench, DGTE wet-crossover bell helper. Experiment 0 ranks #17.

Hard accounting rule

What it is not

It is not a generator and not an energy source. source_credit_kw = 0. It only receives credit if it reduces priming loss, pump-start loss, air ingestion, transfer J/cycle, crossover instability, or shallow-intake failure after added losses and complexity.

Best use: cheap side-bench, then attach only to HTH intake, siphon/thermosiphon, DGTE wet crossover, MGHB coupling, or DGTE‑HTH wet-lock tests if measured helper value is positive.

Bench kit

Materials and sensors for first builds

Water-sealed bell / siphon-primer bench

Materials: tub or tray with water | clear hose/tube | plastic cup, jar, or acrylic bell | outlet lower than water level | dye | measuring jug or scale | stopwatch | transparent debris mesh/filter | rotary/starwheel crossover mockup adapter

Sensors: optional cheap flow meter | low-range differential pressure sensor or manometer | camera for visible bubbles | water-depth ruler | scale or measuring jug | priming timer

Transparent HTH / trompe channel

Materials: clear acrylic tube/channel | replaceable Venturi throat | HDPE/PVC adapters | small separator vessel | needle valves | water reservoir or stream/flume

Sensors: flow meter or velocity probe | differential pressure sensor/manometer | air rotameter or mass-flow sensor | water head ruler | temperature probe | camera for bubble survival

MGHB wave tank cell

Materials: small wave tank or long tote/flume | guided float/piston | load cell or spring scale | linear potentiometer | adjustable damping/load | OWC side chamber

Sensors: ultrasonic or ruler wave gauges | load cell | stroke encoder/potentiometer | DC voltage/current logger | camera | leak detector/visual dye

Pure DGTE rotary/starwheel crossover rig

Materials: 3D printed starwheel | acrylic side plates | O-rings/gaskets | dye tracer | small geared motor or hand crank | two water baths

Sensors: torque proxy/current logger | scale for carry-over | pressure taps | cycle counter | camera | tachometer

Hydraulic balanced lock rig

Materials: two small pressure vessels | check/needle valves | transparent lock chamber | dye tracer | hand pump or syringe | capsule shuttle mockup

Sensors: pressure sensors/manometers | flow meter | current logger if motorized | scale | camera | cycle timer

Integrated DGTE thermal slice

Materials: small hot tank | cold tank | insulation | heat exchanger coil | capsule lane | simple linear generator/coil or force proxy | rotary/lock crossover insert

Sensors: thermocouples | wattmeter for heater | flow meter | pressure taps | current/voltage logger | cycle counter | scale | camera

Decision discipline

Current decision tree

START | |-- Cheapest high-learning test? → Transparent HTH / trompe channel. | If entrainment + pressure + loaded air output pass: advance HTH outdoor/deep-pool test. | If fail: salvage Venturi/degassing data and demote HTH power claim. | |-- Highest-upside coastal branch? → MGHB wave tank cell. | If wave reduction and useful work beat baselines: advance mini-breakwater array. | If power weak but wave reduction strong: salvage as smart breakwater. | |-- Pure DGTE viability? → Starwheel crossover and/or hydraulic balanced lock. | If transfer J/cycle, carry-over, or friction fail: pure DGTE waits. | If pass: integrate best crossover into thermal slice. | |-- Hybrid? → Modular assist slice only after parent measurements. | If assisted mode beats best parent after input + complexity penalty: hybrid branch survives. | If not: keep modular tandem, reject over-integrated claims. | |-- Water-sealed bell? → Experiment 0 side-bench. | If net_helper_value > 0 and repeated cycles survive: attach as helper. | Always source_credit_kw = 0. Never call it a generator. | |-- Solar? → Same-area PV/PV‑T comparison first. | If PV/PV‑T wins: solar DGTE becomes special-case top-up only.