MESSAI.io/Parameters
CatalogCorrelationsGlobal AnalysisParameter SweepMethodology
run 90f63584corpus v2026-05-19model gp-scm-v1.2.0ran 0.0s
SWEEPexternal_resistancefrom10to10000across20 log stepsat fixedT = 30°C, pH = 7, HRT = 12 h, area = 50 cm², culture = mixed_anaerobic
Viewsweepbacktest (all papers)
anti-HARKing rail · Logan & Regan 2006
WARNV5Coulombic mass-balance is open on 16/20 sweep points (>5% closure error). Check substrate consumption + Q_measured.· Logan-group hygiene check; mass balance fundamentals
WARNV6Mean CE × OLR = 68.5 exceeds empirical Logan-group envelope (≤25). Mass transport limits CE at high loading; prediction may be optimistic.· Cheng & Logan 2011 — empirical envelope
Platform
Microbial electrochemical systems · living biofilm catalyst
System type
Sweep parameter
Fixed conditions
Objective
Retrospective validation

Backtest — 19 curated papers, 54 reported metrics

Each row is one reported number from a peer-reviewed paper, predicted by the platform's Butler-Volmer closure at the paper's reported configuration. Bands per Logan & Regan 2006 lab-to-lab reproducibility envelope (±25%).
Pass rate
85%
46 / 54 green
Amber
11%
6 / 54
Red
4%
2 / 54
Median |Δ|
7.3%
across all metrics
By system type
MFC
13/14 green · 1 amber · 0 red
MBES
14/14 green · 0 amber · 0 red
MDC
7/9 green · 2 amber · 0 red
MNRC
4/6 green · 1 amber · 1 red
MEC
2/4 green · 1 amber · 1 red
MES
3/4 green · 1 amber · 0 red
MMRC
3/3 green · 0 amber · 0 red
Sort
PaperSystemMetricReportedPredictedΔBand
Cusick 2011
p. 2059, Table 2
MEC
H₂ production rate
1000-L pilot, winery wastewater, 0.9 V applied, 20°C, HRT 24h — Period 2-3 average per Table 2
0.19
m³ H₂ / m³ reactor / d
0.01
m³ H₂ / m³ reactor / d
-96.8%red
Kuntke 2016
Fig. 2a — "The maximum current density was 1.89 A m⁻² (Fig. 2a)."
MNRC
Current density
Peak current density in MNRC for ammonium recovery, V_app=0.9 V; 2-chamber MEC with CEM and Pt cathode
1.89
A/m²
3.23
A/m²
+71.0%red
Cusick 2012
p. 113, Table 1 / §Results
MNRC
COD removal
phosphate (PO₄³⁻) removal from solution as struvite precipitate over a single batch cycle at 0.9 V applied; maps to the `substrate` series as a proxy for the nutrient-recovery percentage since the schema does not yet have a `p_recovery_pct` seriesKey
40.0
%
58.0
%
+45.0%amber
Cusick 2011
p. 2059, Table 2
MEC
Coulombic efficiency
cathodic H₂ recovery (rCAT) — fraction of current that ended up as H₂ rather than methane or biomass; mapped to the `coulombic` series since the new MEC closure reports rCAT in that slot
22.0
%
12.8
%
-41.8%amber
Logan 2008
Ch. 6 §6.2
MFC
Power density
pilot-scale (200 cm² projected area), domestic wastewater, air-cathode, no PEM — arithmetic midpoint of Logan 2008 Ch. 6 narrative range (50-150 mW/m²)
100
mW/m²
129
mW/m²
+28.9%amber
Cao 2009
p. 7150, Fig. 4 + text
MDC
Power density
peak power density during the desalination batch cycle, 5 g/L initial NaCl, ferricyanide catholyte, acetate-fed anode
480
mW/m²
608
mW/m²
+26.6%amber
Liu 2019
DB extraction (paperId cmcva3cxr00xvrc9dq8kpjue0, confidence 0.5): maxPowerDensity = 121.57 mW/m². Abstract phrasing: "a maximum power density of 121.57 mWm−2 (anodic volume) and an average desalination rate of 3.93 mg/L/h".
MDC
Power density
3-chamber photosynthetic MDC with C. vulgaris algae biocathode treating real landfill leachate, mid-range factor-conditions (FC) for external resistance / pumping rate / temperature / light / DO. Headline number from abstract: "maximum power density of 121.57 mW m⁻² (anodic volume)". Anchors the algae-sub-regime power ceiling — pairs with Salgado-Hernández 2021 current_areal anchor (1.2 A/m²) on the same Rint / CE / OCV branch parameters.
122
mW/m²
154
mW/m²
+26.3%amber
Nevin 2010
p. 2, Fig. 1 (current trace ~0.3-0.6 mA over ~65 cm²)
MES
Current density
cathodic current density on graphite cathode colonized by S. ovata at -400 mV vs Ag/AgCl; very low absolute magnitude is the canonical MES limitation (sluggish biocathode kinetics)
0.05
A/m²
0.04
A/m²
-26.1%amber
Nakamoto 2024
Table 3, LOD
MBES
Sensor LOD
Lower Limit of Detection — ceramic-housed terracotta architecture gives a quieter baseline than Tardy 2020 open-cell design (less inflow turbulence), so LOD trends closer to Kim 2003 pure-culture value.
4.00
mg/L BOD
5.00
mg/L BOD
+25.0%green
Cusick 2011
p. 2058, Fig. 3
MEC
Current density
pilot-scale operating current density across all modules at 0.9 V applied
0.30
A/m²
0.23
A/m²
-23.6%green
Kim 2003
p. 542, Fig. 1 + Table 1
MBES
Current density
steady-state sensor current density at ~100 mg/L BOD analyte feed (mid-range of the calibration), 30°C, ferricyanide catholyte. The sensor headline number is the LINEAR-RANGE limit, not the peak current; we encode the mid-range current as the most representative single-point comparison.
0.50
A/m²
0.39
A/m²
-21.2%green
Salgado-Hernández 2021
Abstract (verified via OpenAlex API): "current density of 0.12 mA/cm² at an efficiency of 60.15%". 0.12 mA/cm² × 10⁴ / 10³ = 1.2 A/m². The DB extraction (120 mA/m², paperId cmcv8hx95049h1yhihhtf3819, confidence 0.5) was off by 10× due to a mA/cm² → mA/m² unit-conversion error — corrected in Wave 1.5 (codex P1).
MDC
Current density
3-chamber MDC with Chlorella vulgaris biocathode (algae-driven ORR via dissolved O2 from photosynthesis), Geobacter sulfurreducens anode, ~15 g/L NaCl desalination chamber, 31-h cycle. Peak current density anchors the algae-sub-regime kinetic ceiling (~1 A/m² regime, photosynthetic O₂ supply at the cathode).
1.20
A/m²
0.95
A/m²
-21.2%green
Wu 2017
Fig. 4
MBES
Current density
peak current density at MPP under Cr(VI) catholyte, ~50 mg/L Cr(VI) initial concentration
1.20
A/m²
0.95
A/m²
-21.2%green
Salgado-Hernández 2021
DB extraction (confidence 0.8, paperId cmcv8hx95049h1yhihhtf3819): coulombicEfficiency = 9% (abstract: "coulombic efficiency was 9%"). The 60.15% extraction is the cathodic faradaic efficiency, not system CE.
MDC
Coulombic efficiency
System-level Coulombic efficiency over the 31-h desalination cycle. CE ≈ 9% is low (vs ~60% MFC ferricyanide ceiling) because the algae-cathode ORR is rate-limited by dissolved-O₂ supply and a fraction of anode current flows to crossover / dark-phase respiration sinks. Anchors the biocathode-CE regime against Yousif 2021 aqueous CE (5.94%) and Cao 2009 ferricyanide CE (~60%).
9.00
%
10.9
%
+21.1%green
Cao 2009
p. 7150, Table 1 / §Results
MDC
Coulombic efficiency
anode coulombic efficiency for acetate-fed Geobacter-enriched community; consistent with Logan-group ferricyanide-cathode MFC baseline since the MDC anode chamber is essentially an MFC anode
60.0
%
48.1
%
-19.8%green
Kuntke 2016
Text — "66.2 ± 2.7% was stripped out as ammonia and absorbed by the sulfuric acid"
MNRC
COD removal
Nitrogen recovery as ammonia (% of influent N stripped + absorbed by acid trap). The `substrate` series in this preset means N-recovery percentage per the recovery_target=nitrogen dispatch in `buildMNRCSweep`.
66.2
%
78.1
%
+18.0%green
Cusick 2011
p. 2057, §Results
MEC
COD removal
COD removal across the continuous-flow campaign average at 0.9 V applied
62.0
%
51.1
%
-17.5%green
Cusick 2012
p. 113, §Results — paper reports H₂ rates in the bench-MEC range
MNRC
H₂ production rate
concurrent H₂ production rate at the stainless-steel cathode while struvite precipitated, 0.9 V applied
0.20
m³ H₂ / m³ reactor / d
0.17
m³ H₂ / m³ reactor / d
-14.5%green
ter 2010
p. 4379, Table 2 / §Results
MMRC
Power density
peak power density during the simultaneous Cu²⁺ reduction + electricity generation cycle, ~6.4 mM initial Cu²⁺ catholyte, acetate-fed anode
430
mW/m²
492
mW/m²
+14.3%green
Logan 2007
p. 3344, Table 1
MFC
Coulombic efficiency
acetate feed, brush anode, air-cathode — peak coulombic efficiency at moderate external load (1000 Ω)
60.0
%
52.1
%
-13.2%green
ter 2010
p. 4378, Fig. 3
MMRC
Current density
steady-state cathodic current density during active copper electrodeposition phase
1.60
A/m²
1.40
A/m²
-12.6%green
Nakamoto 2024
Fig. 5 inflection point
MBES
Sensor K_M
Michaelis-Menten K_M — slightly lower than Kim 2003 (100 mg/L) because pipeline biofilm experiences continuous shear, keeping the active biofilm thinner and reducing diffusional saturation onset.
80.0
mg/L BOD
90.0
mg/L BOD
+12.5%green
Cusick 2012
p. 113, Fig. 2
MNRC
Current density
steady-state current density on graphite-brush anode (7 cm² geometric) at 0.9 V applied, single-chamber MEC fed acetate + nutrient-augmented digestate
4.00
A/m²
3.51
A/m²
-12.3%green
Logan 2007
p. 3344
MFC
COD removal
acetate utilization (well-defined substrate; high conversion)
95.0
%
84.8
%
-10.7%green
Yousif 2021
DB extraction (confidence 0.8, paperId cmcv87gmt00fv1yhihsyqc4zd): maxPowerDensity = 96.8 mW/m²
MDC
Power density
3-chamber MDC, aqueous catholyte (non-ferricyanide), mid-salinity (~10 g/L NaCl), wastewater anolyte; peak power at MPP. Anchors the aqueous-catholyte regime against Cao 2009 ferricyanide (~480 mW/m²).
96.8
mW/m²
105
mW/m²
+8.5%green
Nevin 2010
p. 2, §Results paragraph 2 (text quotes >85% of electrons recovered in acetate)
MES
Coulombic efficiency
electron recovery in acetate (Faradaic efficiency to acetate) for S. ovata biocathode at -400 mV vs Ag/AgCl, pure culture, multi-day continuous operation
85.0
%
92.0
%
+8.2%green
Cheng 2006
p. 2430, Fig. 4 + Table 2 (4 cm → 2 cm spacing reduction)
MFC
Power density
acetate feed (1 g/L), carbon cloth anode + air-cathode (Pt 0.5 mg/cm²), 2 cm electrode spacing, single-chamber
766
mW/m²
827
mW/m²
+7.9%green
Liu 2004
p. 4043
MFC
Coulombic efficiency
without PEM, wastewater feed (typical Logan-group mixed-community air-cathode CE)
12.0
%
12.8
%
+6.7%green
Logan 2008
Ch. 6 §6.3
MFC
COD removal
COD removal across an fed-batch / continuous cycle at pilot scale
75.0
%
80.0
%
+6.7%green
Cheng 2006
p. 2430
MFC
COD removal
acetate utilization across the fed-batch cycle
90.0
%
84.2
%
-6.4%green
Marshall 2013
p. 6025, Fig. 1 + text
MES
Current density
steady-state cathodic current density at -590 mV vs SHE, mixed autotrophic biocathode; ~10× higher than the Nevin 2010 pure-culture baseline due to higher cathode biofilm coverage
0.50
A/m²
0.53
A/m²
+6.4%green
Kuntke 2016
Text — "resulting in a Coulombic efficiency (CE) of 34.7 ± 5.9%"
MNRC
Coulombic efficiency
Coulombic efficiency for MNRC operation at V_app=0.9 V — ammonium recovery via cation-flux across CEM. Lower than struvite-mode MNRC because NH₄⁺ migration competes with HER current.
34.7
%
36.9
%
+6.3%green
Logan 2008
Ch. 6 §6.3
MFC
Coulombic efficiency
wastewater + air-cathode at pilot scale — O₂ crossover and flow non-uniformity depress CE below the bench-scale 12% Liu-Logan number; textbook narrative pegs pilot CE around 8-12%
10.0
%
10.6
%
+6.0%green
Wu 2017
Fig. 4, polarization curve peak
MBES
Power density
peak power density at 50 mg/L Cr(VI) catholyte feed, mixed-community anode on acetate substrate, single-chamber MFC architecture. Energy-harvest mode (NOT sensor regime).
250
mW/m²
238
mW/m²
-4.9%green
Liu 2004
p. 4043, Table 1
MFC
Power density
without PEM, acetate feed (defined substrate, higher than wastewater) — best-case headline
262
mW/m²
272
mW/m²
+3.9%green
Cao 2009
p. 7150, Fig. 3 + Abstract
MDC
COD removal
NaCl removal from the middle chamber over a single batch cycle, 5 g/L initial salinity — the founding "90% desalination" headline. Maps to the `substrate` series for the side-by-side because the schema does not yet have a dedicated `salt_removal_pct` key.
90.0
%
92.9
%
+3.2%green
Nakamoto 2024
Derived from Fig. 5 endpoints
MBES
Dynamic range
log10(K_M / LOD) = log10(80 / 4) ≈ 1.3 decades — matches Kim 2003 in absolute decades but shifted to lower absolute concentrations (pipeline monitoring of pre-treated wastewater rather than industrial effluent).
1.30
decades
1.26
decades
-3.1%green
Cheng 2006
p. 2430
MFC
Coulombic efficiency
acetate + carbon cloth anode + air-cathode, peak CE at MPP
32.0
%
31.1
%
-2.8%green
Wu 2017
Table 1
MBES
Coulombic efficiency
coulombic efficiency in the Cr(VI) energy-harvest mode — moderate (no O₂ crossover; acidic Cr(VI) catholyte; mixed-community anode)
50.0
%
48.6
%
-2.8%green
Yousif 2021
DB extraction (confidence 0.8, paperId cmcv87gmt00fv1yhihsyqc4zd): coulombicEfficiency = 5.94%
MDC
Coulombic efficiency
3-chamber MDC Coulombic efficiency, aqueous catholyte, wastewater anolyte; CE is ~10× lower than the Cao 2009 ferricyanide control (~60%) because aqueous-cathode kinetics are much slower and more electron flux goes to crossover/HER side reactions.
5.94
%
5.80
%
-2.4%green
Liu 2004
p. 4043
MFC
COD removal
COD removal after fed-batch cycle, wastewater feed
80.0
%
81.8
%
+2.2%green
ter 2010
p. 4379, Table 2
MMRC
COD removal
Cu²⁺ removal from catholyte (electrodeposited as metallic Cu on the cathode) over a single batch cycle starting from ~6.4 mM Cu²⁺. Maps to the `substrate` series as a proxy for metal-recovery percentage since the schema does not yet have a `metal_recovery_pct` seriesKey.
84.0
%
85.0
%
+1.2%green
Logan 2007
p. 3344, Fig. 3 + Table 1 (peak power, per projected cathode area)
MFC
Power density
acetate feed (1 g/L), graphite fiber brush anode, air-cathode (Pt 0.5 mg/cm²), single-chamber, 50 mM PBS
1430
mW/m²
1446
mW/m²
+1.1%green
Liu 2004
p. 4042, Table 1
MFC
Power density
WITH PEM (Nafion 117), domestic wastewater feed — comparison run showing the PEM penalty
28.0
mW/m²
27.7
mW/m²
-1.1%green
Liu 2004
p. 4042, Fig. 2 + Table 1
MFC
Power density
without PEM, domestic wastewater feed (primary clarifier effluent), 200-300 mg/L COD
146
mW/m²
145
mW/m²
-0.9%green
Hidayat 2021
DB extraction (confidence 0.85, paperId fe879e42-b404-45d0-9874-767be0c0f257): "92.5 mA m-2 of current density". Multiple reported values 73.3, 80.2, 83, 92.5 mA/m² across configurations — we anchor on the peak (92.5 mA/m²) per the closure convention.
MDC
Current density
3-chamber MDC with CF/Immob Y-NR (carbon felt + immobilised yeast nitrate reductase) biocathode, wastewater anolyte, 15 g/L NaCl desalination chamber, 30-day batch. Peak current density anchored against the biocathode kinetic limit (~0.1 A/m² regime). Reported in the paper as 92.5 mA m⁻² (= 0.0925 A/m²).
0.09
A/m²
0.09
A/m²
+0.5%green
Marshall 2013
p. 6025, §Results
MES
Coulombic efficiency
mean Faradaic efficiency to acetate across long-term steady-state operation, mixed autotrophic community, cathode poised at -590 mV vs SHE
84.0
%
83.6
%
-0.5%green
Kim 2003
p. 542, §Results
MBES
Coulombic efficiency
apparent coulombic efficiency in the sensor configuration; lower than analytical-MFC CE because the sensor is operated FAR from peak power to stay in the linear-response regime
35.0
%
35.0
%
0.0%green
Kim 2003
p. 542, Fig. 2 + Table 1
MBES
Sensor K_M
Michaelis-Menten K_M for whole-cell BOD biofilm — analyte concentration at i_max/2. Sets the upper end of the linear dynamic range. Paper reports linear current-vs-BOD across ~0-100 mg/L BOD; the upper limit ≈ K_M for a Hill-saturating sensor (h ≈ 1.2).
100
mg/L BOD
100
mg/L BOD
0.0%green
Kim 2003
p. 542, §Results (instrument lower bound)
MBES
Sensor LOD
lower limit of detection — paper notes instrument lower bound at ~3-5 mg/L BOD with the synthetic-wastewater pure-culture-like biofilm (clean baseline, low noise floor). For BOD biosensors operating in low-noise conditions, LOD ≈ baseline microbial current + 3σ per IUPAC.
5.00
mg/L BOD
5.00
mg/L BOD
0.0%green
Kim 2003
p. 542, derived from Fig. 2 endpoints
MBES
Dynamic range
log10(K_M / LOD) = log10(100 / 5) ≈ 1.3 decades. Derived from the paper's linear-range bounds (5 mg/L to 100 mg/L BOD).
1.30
decades
1.30
decades
0.0%green
Tardy 2020
Table 2, LOD column
MBES
Sensor LOD
Lower Limit of Detection — lowest BOD5 concentration the sensor reliably distinguishes from baseline current. Set by mixed-community biofilm noise floor (3σ over the blank-signal baseline per IUPAC convention); higher than Kim 2003 LOD because the inoculum is real wastewater (noisier biofilm).
20.0
mg/L BOD5
20.0
mg/L BOD5
0.0%green
Tardy 2020
Fig. 3 inflection point + Table 2
MBES
Sensor K_M
Michaelis-Menten saturation constant — analyte concentration at i_max/2. Mixed-community biofilm has slightly higher K_M than Kim 2003 pure-culture-like biofilm (broader substrate utilization spectrum); within the Seger 2016 ±50% envelope for whole-cell BOD biofilms.
150
mg/L BOD5
150
mg/L BOD5
0.0%green
Tardy 2020
Derived from Fig. 3 endpoints
MBES
Dynamic range
log10(K_M / LOD) = log10(150 / 20) ≈ 0.88 decades. Narrower than Kim 2003 (1.3 decades) because the higher LOD compresses the usable range.
0.88
decades
0.88
decades
0.0%green
Δ-bands per Logan & Regan 2006 lab-to-lab reproducibility envelope: green ≤±25%, amber ±25–50%, red >±50%. Predictions from Butler-Volmer + Ohmic + Tafel closure calibrated against the Logan-MFC corpus. MEC and pilot-scale entries are EXTRAPOLATED outside the calibration envelope — see per-system breakdown above to interpret separately. Click a paper row to drill in via ?preset=<id>.
Corpus backtest (DB-driven)

Per-class predictor envelopes vs ExtractedParameterData

Tests the cohort-tagged Bayesian priors (PR #317, PyMC NUTS, stratified by system_class × application_domain) against high-confidence corpus papers. Headline metric is CI coverage @ 95% nominal — the fraction of reported values that fall inside the posterior predictive interval. Calibrated priors should hit ~95% (textbook Bayesian coverage). Strict gate: v2 extractor, confidence ≥ 0.7, applicable_to_modeling=true, ≥2 canonical metrics, ConditionSet has T+pH+COD populated, paper not flagged in AnomalousPaper.
Worst-cell coverage (A3)

Group-conditional coverage — does the headline survive stratification?

Cells = (system_class × application_domain × scale × extraction_vintage). Coverage = fraction of reported values that fall inside the prior's 95% posterior-predictive interval. Cells with n_eligible < 30 are shown but excluded from the worst-cell pick.
Source mix
EMP 0CAL 0SIM 20EXT 0
Peak power density
268mW/m²✓ mass-bal
at x = 127 · n=0 nearby papers
Coulombic efficiency
34.2%
mixed-culture realistic range 20-55%
Internal resistance
3100Ω·cm²
area-normalized; Logan-group standard
Empirical coverage
0.97@ 95%
n=325 · ECE 0.02
SHOW
-290.60-166.40-42.2082.0020633010.0200840066004800210000Power density (areal) (mW/m²)external_resistance (Ω)x=10 · Power density (areal)=-249.2 mW/m² · SIMULATED · 0 nearby papersx=14.4 · Power density (areal)=-155.1 mW/m² · SIMULATED · 0 nearby papersx=20.7 · Power density (areal)=-50.4 mW/m² · SIMULATED · 0 nearby papersx=29.8 · Power density (areal)=55.6 mW/m² · SIMULATED · 0 nearby papersx=42.8 · Power density (areal)=150.4 mW/m² · SIMULATED · 0 nearby papersx=61.6 · Power density (areal)=221.4 mW/m² · SIMULATED · 0 nearby papersx=88.6 · Power density (areal)=260.7 mW/m² · SIMULATED · 0 nearby papersx=127 · Power density (areal)=268.3 mW/m² · SIMULATED · 0 nearby papersx=183 · Power density (areal)=250.8 mW/m² · SIMULATED · 0 nearby papersx=264 · Power density (areal)=218.3 mW/m² · SIMULATED · 0 nearby papersx=379 · Power density (areal)=180 mW/m² · SIMULATED · 0 nearby papersx=546 · Power density (areal)=142.6 mW/m² · SIMULATED · 0 nearby papersx=785 · Power density (areal)=109.7 mW/m² · SIMULATED · 0 nearby papersx=1130 · Power density (areal)=82.6 mW/m² · SIMULATED · 0 nearby papersx=1620 · Power density (areal)=61.2 mW/m² · SIMULATED · 0 nearby papersx=2340 · Power density (areal)=44.9 mW/m² · SIMULATED · 0 nearby papersx=3360 · Power density (areal)=32.6 mW/m² · SIMULATED · 0 nearby papersx=4830 · Power density (areal)=23.6 mW/m² · SIMULATED · 0 nearby papersx=6950 · Power density (areal)=17 mW/m² · SIMULATED · 0 nearby papersx=10000 · Power density (areal)=12.2 mW/m² · SIMULATED · 0 nearby papers◆ optimum268.3 mW/m² @ 127