SCROLL TO FOLLOW THE FLOWPUMP → RAIL → ENGINE → RETURN
ONE SYSTEM. THREE FLOWS.02 — THE BALANCE
EVERY LITRE HAS A JOB.
Some fuels the engine. The rest must return to the tank while the regulator holds pressure. See the whole balance, from a quiet idle to your chosen boost level.
01 / IN
RAIL SUPPLY
Fuel reaching the rail at the operating pressure.
−
02 / USED
ENGINE DEMAND
Fuel the engine consumes at your chosen load.
=
03 / BACK
RETURN REQUIRED
Surplus the regulator and return path need to pass.
THE BARS ILLUSTRATE THE PRINCIPLE, NOT A MEASURED RESULT.NEXT: YOUR SETUP ↓
YOUR PARTS. YOUR OPERATING POINT.03 — THE COMPARISON
BUILD IT. BALANCE IT. KNOW THE FLOW.
Start with ten Evo X configurations. Mix and match the pump, rail and regulator. Then change the fuel, boost and engine demand to see how the estimates respond.
Interactive estimates, with assumptions and source data included.
CHANGE AN INPUT. SEE THE SYSTEM RESPOND.THE FULL REPORT IS NEXT ↓
EVO X / FUEL SYSTEMS
4B11T · Engineering comparison · 9 September 2026
Pump + rail + regulator
Fuel in. Fuel used. Fuel returned.
Ten mix-and-match Evo X configurations, calculated at your fuel pressure and engine demand. All flow results are estimates at the target pressure.
Read the numbers correctly: return flow is rail supply minus engine use, assuming the FPR can hold the target pressure. An upgraded rail or FPR does not add a fixed number of L/h. The OEM pump curve is unavailable; enter measured data to calculate the all-stock baseline. The AEM 320 option is the E85-compatible 50-1200, using a reconstructed legacy curve.
Selected combination
02 · AEM 320 (legacy) / Stock / Stock FPR
Available at the rail247.9 L/hReconstructed manufacturer graph
Engine fuel demand164.4 L/h450 crank hp · BSFC 0.60
Return flow required83.5 L/hRequired bypass, if target pressure is held
Margin available. The FPR and return path must pass 83.5 L/h at load and 316.6 L/h in the idle scenario. Bypass capability remains conditional on the complete installed return path. AEM results use a reconstructed legacy 320 curve. This operating point is close to the 80 psi data boundary; verify the pump revision and relief behavior.
Reference pump output247.9 L/h
Target rail / pump pressure73.0 / 76.4 psi
Internal jet diversion0.0 L/h
Supply margin / engine demand+50.8%
Idle return required316.6 L/h
38.0 psi rail · 13.5 V · 2.0 L/h engine use. FPR + return capacity is unverified.
Pump-side power ceiling with reserve617 hp at crank
10% flow reserve, current boost and BSFC. Pump supply only; injectors and engine limits are unverified.
Compare all 10 combinations
Petrol · 450 crank hp · 25.0 psi boost · 13.5 V · 10% flow reserve target
Select a build number to edit
Build
Pump
Rail / FPR
Rail supply L/h
Engine L/h
Return L/h
Idle return L/h
Supply result
Stock OEM
Curve not available
Stock railStock FPR
—
164.4
—
—
OEM pump curve required
AEM 320 (legacy)
Legacy graph estimate
Stock railStock FPR
247.9
164.4
83.5
316.6
Margin available
AEM 320 (legacy)
Legacy graph estimate
AMS Evo X railStock FPR
247.9
164.4
83.5
316.6
Margin available
AEM 320 (legacy)
Legacy graph estimate
Stock railAEM 25-302BK
247.9
164.4
83.5
316.6
Margin available
AEM 320 (legacy)
Legacy graph estimate
Radium Evo X railRadium DMR-RA
247.9
164.4
83.5
316.6
Margin available
DW65c
Manufacturer table
Stock railStock FPR
214.4
164.4
50.1
260.3
Margin available
DW300c
Manufacturer table
AMS Evo X railAEM 25-302BK
264.7
164.4
100.3
329.8
Margin available
450 · F90000267
Manufacturer table
Radium Evo X railAEM Hi-Cap 25-305BK
330.5
164.4
166.1
409.5
Margin available
450 HP · F90000274
Manufacturer table
Stock railAEM Hi-Cap 25-305BK
342.6
164.4
178.2
412.7
Margin available
450 HP · F90000274
Manufacturer table
Radium Evo X railRadium DMR-RA
322.3
164.4
157.9
396.2
Margin available
All rows share the operating-point inputs. The first nine start with an OEM-style feed/return arrangement; build 10 includes a current Radium hanger and revised plumbing. Every pressure-loss value is an editable assumption. “Margin available” assesses supply only; it does not certify FPR bypass capacity or fitment.
Mix and match the selected build
Change parts independently. Matching the hardware does not automatically change the assumed hydraulic losses.
Pump: Legacy 320-E series, 13.5 V. Digitized graph shape normalized to 320 L/h at 43 psi. Not the gasoline 50-1000 or a substitute for every current 340 revision. Needs an appropriate Evo X installation kit or matched hanger. Source
Rail: Retains the factory rail. Pressure loss is not known from a verified installed test. Source
FPR: Uses the separately entered measured OEM base pressure. No assumed L/h bypass limit. Source
OEM return hose and jet remain a possible pressure-control restriction. Replacing the rail/FPR alone does not remove them.
Idle bypass capacity is unknown. A regulator horsepower label is not a maximum return-flow rating.
OEM jet flow is return-driven: it is not deducted from rail supply. Current Radium hangers use a supply diversion; 30 L/h is the starting assumption within Radium’s published 15–45 L/h range. Older hanger revisions can differ. Loss coefficients and jet flow are not automatic part ratings.
What happens as boost rises?
Selected build. Engine power/duty and all other inputs stay fixed during this pressure sensitivity sweep.
Rail supplyReturn requiredEngine demand
25.0 psi → rail 247.9 L/h · return 83.5 L/h
Lines stop where reference data stops. Missing curve segments are not zero flow. Pump losses are solved again at each boost point; return lines assume the FPR can regulate.
How to use this result
Flow capacity and pressure regulation are different checks.
1. Size at full fuel pressure
The pump must overcome base pressure, boost and feed losses. A pump’s “320” or “450” name is not its delivery at 25 psi boost. A rail upgrade can reduce a restriction, but this report does not multiply pump flow by marketing claims.
2. Check the return at idle
Low engine consumption leaves more fuel for the FPR to bypass. The Evo X return hose and in-tank jet can raise return backpressure. An adjustable FPR cannot necessarily correct a restriction farther downstream. Radium’s Evo X notes.
3. Read “shortfall” literally
If demand exceeds capacity, the target pressure cannot be maintained. The displayed shortfall is at that target pressure; the report does not predict the resulting lower rail pressure, AFR or engine output. A zero-return result is not adequate sizing headroom.
4. Confirm on the actual car
Use measured rail pressure minus manifold pressure, pump-terminal voltage and fuel consumption at load. Confirm low-load pressure and return behavior too. The report does not establish injector capacity, compatible fittings, filter choice or a complete installation bill of materials.
Pressures are gauge psi, flow is L/h. Both K values are user-assumed pressure losses at 300 L/h. Zero tank pressure and negligible static head are assumed. The rail is treated as one pressure node; injector-to-injector pulsation and pressure gradients are not resolved.
The solver finds the intersection of the pump curve and the feed-loss curve. Internal supply diversion recirculates in the tank; it is not engine consumption or FPR return. Actual FPR return equals this result only when the target pressure is held.
Qengine = crank hp × BSFC × 0.45359237 / density crank hp = wheel hp / (1 − drivetrain loss)
1 hp = 0.745699872 kW. Density is kg/L and BSFC is lb/(hp·h). Preset density and BSFC are editable assumptions. E85 composition, temperature and tune vary. Fuel selection changes demand assumptions, not the manufacturer pump curve’s test fluid.
Duty is a fraction in this equation. Injector ratings must match the fuel being evaluated. The square-root correction assumes the FPR holds constant differential pressure. In injector mode the calculated demand can change with the chosen regulator’s base pressure.
The factory regulator’s nominal 329 kPa (48 psi) pressure is the starting base value. Replace it with the measured reference-off value for your car. Manufacturer tables are interpolated linearly in pressure. TI data is also interpolated between 12 and 13.5 V. Other voltage values use the optional approximation (enabled by default): Q(V) = Q(Vref) × [1 + slope × (V − Vref)]. It is a sensitivity assumption, not a motor model. The optional derating is then applied once. No pressure extrapolation is displayed.
The reserve-adjusted power ceiling divides available rail flow by (1 + reserve) before converting it to engine power using the selected fuel and BSFC. It is not a validated horsepower rating for the car. The idle check uses its own pressure, voltage and measured/assumed consumption.
The measured idle bypass limit is checked only at the recorded idle rail pressure and fuel density. Exceeding it flags a likely pressure-control failure; the report does not cap the calculated return and pretend pressure remains correct. Unknown bypass capacity stays unknown.
Pump data, exact points & evidence quality
AEM 320: legacy 50-1200/50-1205 series data, not a substitute for every pump sold as AEM 320/340. The gasoline 50-1000 is not modeled. The 2015 AEM chart’s gasoline 0.5-BSFC series was digitized, then its relative shape normalized to 320 L/h at 43 psi. This reconstruction avoids assuming an undocumented test-fluid density. The plotted pressure range is 35–80 psi at 13.5 V. The AEM test included its filter and FPR, so it is an assembly-curve proxy; added losses represent further assumed installation losses. Display rounding does not imply measurement precision.
The current AEM product page says 340 L/h at 40 psi and lists an 80 psi relief point; its linked revised manual says 340 L/h at 43 psi. Those differing ratings are not merged into the legacy curve. Confirm the exact revision and curve for your pump.
DW65c and DW300c: numeric tables from DeatschWerks’ linked manufacturer sheets, revision 12 October 2023, at 13.5 V. The DW300c is not the larger DW300; their curves differ.
TI/Walbro: exact F90000267 and F90000274 manufacturer tables. Test fluids and dates differ from the DW and AEM data. Small numerical differences are not a controlled cross-brand comparison. F90000274 is the higher-relief-pressure version; do not interchange these curves.
OEM: no flow curve is supplied. All-stock means the three named parts, not a complete factory vehicle calibration. User-supplied pump data is explicitly identified. FPR bypass capacities, installed feed losses and rail-specific pressure drops were not established from the retrieved sources.
AMS.04.07.0006 series, includes a feed line and pulsation damper. Factory FPR can be retained. Aftermarket FPRs need the correct rail adapter and plumbing.
Direct mounting suits a compatible 8AN ORB port. Other rails need a compatible adapter/remote arrangement; check clearance. 150 psi is a conservative report scenario ceiling, not a manufacturer DMR maximum.
All combinations are planning configurations. Use a vehicle-specific pump installation kit or a hanger explicitly matched to the chosen pump. Universal pumps, universal FPRs and arbitrary rail/FPR pairs do not become direct-fit parts simply because they can be selected here. When replacing the OEM module, provide the required downstream filtration. Match fuel compatibility across hoses, seals, filter, injectors and regulator.
Sources & audit trail
Primary manufacturer documents and product information retrieved 9 September 2026. AEM’s historical PDF is manufacturer-authored and hosted by a distributor. Links are for evidence; the calculator itself runs offline.
Manufacturer service manual hosted by an Evo X manual archive. Nominal regulator pressure: 329 kPa (48 psi); test range 45–50 psi with reference disconnected and about 38 psi connected at idle. Use your measured base pressure.
21 May 2019; Stoddard test fluid; 12 and 13.5 V. The model uses the unambiguous 30–100 psi tabulated points, including the marked high-pressure flow drop.
15 January 2019; Solvent 142-66; 12 and 13.5 V. US gallons/hour converted with 3.785411784 L/US gal. Ambiguous repeated low-pressure samples are omitted; conservative model overlap is 38–110 psi.
Factory-line/FPR compatibility, included feed line and pulsation damper. Flow-area and marketing improvement claims are not applied as whole-system multipliers.
OEM return/jet restrictions; current jet consumes 15–45 L/h. The 8AN feed and OEM-feed-as-return conversion are described. A current hanger is explicitly distinguished from earlier designs.