Second River Pilot
Build Specification — 3 Modules

Why This Site Is Easy

Most HVEH sites need an engineered basin: a concrete channel built to force stormwater into a tangential intake and hold the initial radius above the attractor threshold r* ≈ 0.776. The Second River / Watsessing confluence already does this for free. Two flows meeting at a natural junction provide the rotational component the helical attractor requires — no channel construction needed, just a guide-vane structure to direct the confluence entry and a rotor to sit in the vortex axis.

What You're Actually Building

Three identical modular installations, one at the confluence and two upstream on the Second River reach. Each module has four physical parts:

1 · Guide Vane Structure (the K element)

A low-profile concrete or steel deflector — essentially a shaped sill — set at the confluence entry point. Its job is to angle the two incoming flows into a tangential spiral entry rather than a direct head-on collision. This is the K operator (curvature gate): it must be installed and stable before the turbine is lowered. Dimensions: roughly 2–4 m span across the channel, 0.3–0.6 m above low-water level. No excavation; anchored to existing channel bed.

2 · Vertical-Axis Rotor

A custom helical rotor, ~0.8–1.2 m diameter, suspended vertically at the vortex axis on a mild-steel A-frame above water level. The blades are swept to match the contact-geometric spiral — this is the novel piece and the longest-lead item. Blade pitch and sweep angle are derived from the channel geometry measured in the hydraulic survey (§3). Not an off-shelf turbine: needs to be fabricated.

Target flow range for the Second River: 0.3–1.0 m/s normal flow, up to 2.5 m/s storm peak. Rotor design must extract useful torque across this range — wider than a typical micro-hydro turbine optimised for a single design flow.

3 · Generator + Frame

Permanent magnet generator (PMG), ~15–25 kW rated, mounted above water on the A-frame. Direct-drive preferred (no gearbox = lower maintenance). Off-shelf low-head PMG class — several manufacturers (e.g. Ginlong/Solis, Enercon small-wind class) produce suitable units that can be adapted. The generator sits dry; only the rotor enters the water.

4 · Controls + Grid Tie

Off-shelf PLC control cabinet with K-before-F interlock (software prevents rotor spin-up until K-element confirmation signal). Grid tie inverter + disconnect. Cable to nearest Belleville utility distribution point — the Second River corridor has JCP&L distribution infrastructure within ~200 m of the confluence. Net metering agreement with JCP&L for sub-100 kW.

Step 0 — Hydraulic Survey (First)

Before any design work, you need measured flow data at the specific confluence point. Two weeks of field work with an ADCP (acoustic Doppler current profiler) — rentable — plus a total station for channel geometry. This determines:

• Actual flow velocity at low, normal, and storm conditions
• Confluence angle and rotational component already present
• Whether r(0) is above r* ≈ 0.776 at normal flow — confirming mill mode viability
• Channel cross-section for rotor sizing
• Sediment load estimate (Second River drains a developed basin — expect some sediment management)

If the StreamStats data confirms the flow range, field survey confirms the geometry, and r(0) > r* checks out — you move to rotor design. If r(0) is marginal at low flow, the guide vane can be adjusted to compensate (smaller entry angle, tighter spiral). If r(0) is consistently below r* even at normal flow, this site is storm-event-only, not continuous-mode.

Build Sequence

Cost Breakdown — 3 Units

The cost reduction relative to the Chapter 9 estimate (~$450k for 3 medium units) comes entirely from eliminating the engineered concrete basin — the confluence does that work for free. The civil cost drops from ~$55k per unit (full basin) to ~$6k (guide vane only). Everything else is the same.

Permit Checklist

• USGS StreamStats pre-screen — free, Day 1
• NJ GeoWeb wetlands check — confirm if confluence is in FW1/FW2 buffer or upland; determines permit type (Letter of Interpretation vs. Individual Permit)
• NJDEP Flood Hazard Area Individual Permit — N.J.A.C. 7:13; primary permit; ~$800 fee; 6–9 month typical timeline
• NJDEP Freshwater Wetlands — Letter of Interpretation first; Individual Permit if within 150 ft buffer
• NJDEP Division of Fish and Wildlife — coordination letter; fish screen specification for Second River (American eel, potential anadromous species)
• Belleville DPW — stream access permit for in-channel K-element installation; construction permit for above-grade A-frame and electrical
• JCP&L net metering application — sub-100 kW residential/commercial; standard form; 30–60 day review
• No FERC license required — non-navigable waterway, below 100 kW per unit
• No USACE Section 404 — no dredge or fill of navigable waters (guide vane is above bed; confirm with site survey)
• Essex County — confirm no county easement on Second River corridor at Watsessing; deed search
• Belleville sewer authority — Second River corridor may have combined sewer infrastructure nearby; confirm no conflicts with CSO outfall locations

Timeline — Survey to Power-On

The One Hard Thing

Everything in this build is standard procurement except the rotor. The guide vane is a shaped concrete deflector — any civil contractor can build it. The A-frame is mild-steel fabrication. The generator is off-shelf. The permitting is the NJDEP small hydro lane, which exists and is well-defined.

The rotor is novel. A helical vortex turbine that extracts rotational kinetic energy — not axial flow — at 0.3–1.0 m/s in a 3–6 m wide channel does not exist as a catalogue item. The blade geometry must be derived from the contact-geometric analysis (the same analysis that produced r* ≈ 0.776 and the K-before-F protocol) and validated by CFD before fabrication. This takes 3–4 months and is the critical path item that determines the whole timeline.

Route to rotor design

Three options, in order of cost and speed:

1. University partnership — NJIT, Rutgers, or Stevens Institute of Technology hydraulic engineering programmes. Graduate student + faculty supervisor. Cost: ~$8–15k in research support, 4–6 month timeline. Produces peer-reviewable results. Best option for the first unit; strengthens grant applications.
2. Hydraulic engineering consulting firm — e.g. Stantec, Kleinfelder, or a boutique turbomachinery firm. Cost: ~$18–25k for CFD study and fabrication drawings. Faster (2–3 months) but more expensive. Best option if grant timeline is tight.
3. In-house with OpenFOAM — Open-source CFD, no license cost. Requires someone with CFD experience. Cost: staff time only, but slower and higher risk if not an existing capability. Only viable if G6 LLC has or hires a CFD engineer.

What You Need Right Now — Next Three Actions

Rotor Design Specification — Contact-Geometric Blueprint

The helical rotor is the novel piece. Every other component in this build — guide vanes, A-frame, PMG generator, control cabinet — is standard procurement. Only the rotor blade geometry does not exist in any catalogue. This section specifies it completely: the mathematics, the dimensional parameters, the CFD pass criteria, and the CAM output format. A Python script (rotor_geometry.py) generates the 3D surface point cloud directly from these equations.

Phase 1 · Contact-Geometric Blade Profile

Coordinate system. Normalised cylindrical coordinates (r, θ, z), r ∈ [0, 1], z ∈ ℝ. The vortex outer radius R_outer = 0.60 m (rotor diameter 1.2 m). Physical radius ρ = r × R_outer.

Zoning. The rotor has two zones separated by the attractor boundary r* ≈ 0.776:

• Core zone 0 ≤ r < r* — no blades. Free slip. The Rankine vortex core and the Reeb orbit live here; any blade inside this radius disrupts the vortex coherence that drives generation.
• Active extraction zone r* ≤ r ≤ 1.0 — rotor blade span. Rotational kinetic energy is extracted here via torque T_z.

Pitch equation — the core result. The contact form on the vortex manifold is α = dz − r²dθ. A blade surface that annihilates α at every point (i.e., lies in the kernel of α) must satisfy:

This pitch range — 37.8° at the inner blade edge rising to 45° at the tip — is the invariant condition. Blades cut to any other pitch angle break the contact alignment and introduce destructive axial lift or turbulence in the core zone. The equations are exact; CFD validates them, not adjusts them.

Chord scaling. C(r) = C_tip × R_outer / r. Chord widens inward proportional to 1/r, maintaining approximately constant spanwise solidity. At tip: C = 60 mm. At r*: C ≈ 77 mm.

Axial scale. Integrating dz = r²dθ from θ = 0 to θ_sweep:

Profile. NACA 0012 symmetric (zero camber, 12% thickness ratio). Symmetric profile essential — any camber introduces an axial lift component that pushes flow along the vortex axis, degrading the rotational extraction and loading the thrust bearing asymmetrically.

Blade count: 3 blades, 120° stagger. Two blades produce torque ripple at 10–30 RPM in the 0.3–1.0 m/s flow range. Five blades increase blockage in the extraction zone and risk core breakdown at low flow. Three blades with helical stagger produce smooth torque distribution and reliable self-starting from rest as flow rises from zero.

Phase 2 · CFD Validation — Pass Criteria

CFD does not adjust the blade geometry — the contact-geometric equations are exact. CFD validates that the blade, as designed, satisfies three physical conditions at the Second River flow range before fabrication drawings are issued.

Pass gate 1 — torque dominates: T_z · ω / (F_z · U_z) ≫ 1. The rotor must produce substantially more rotational power than axial thrust power. If this ratio approaches 1, camber or pitch is wrong.

Pass gate 2 — energy extracted in active zone: Tangential velocity V_θ must collapse toward zero in the annular region 0.776 ≤ r ≤ 1.0 downstream of the rotor. Non-zero residual V_θ here means the blade is not extracting the rotational energy (profile misalignment or insufficient blade count).

Pass gate 3 — core coherence preserved: No vortex core breakdown at r < 0.776. The Rankine core must remain coherent through the rotor disk. Core breakdown at low flow (< 0.3 m/s) is acceptable — that is the cut-in condition.

Phase 3 · Hardware and Fabrication

Material. Primary candidates are 5083 aluminium alloy (marine grade, corrosion resistant, 2.66 g/cm³, easily machined) or carbon-fibre reinforced polymer (CFRP, lower density, superior fatigue life but higher unit cost). For the first three units, 5083 aluminium is recommended — machine shops in the Newark/Jersey City corridor have 5-axis capability for this alloy, and repair is simpler than CFRP in the field.

Structural. No central shaft through the core zone — a central cantilever would require the hub to penetrate r < r*, disrupting the Rankine core. Two options: (a) outer shroud ring connecting blade tips to the shaft above water, or (b) neutral-profile struts connecting at 0.85 r, positioned at blade trailing edges to minimise flow interference. Option (a) is preferred for the first prototype as it simplifies balancing and bearing design.

CAM output. The rotor_geometry.py script generates a 3D point cloud (CSV format, ~330 k surface points for the full 3-blade assembly at the default resolution). Import into FreeCAD → Surface Workbench → BSpline Surface to fit a G³-continuous NURBS surface per blade. Export as STEP for 5-axis CAM. The NURBS fitting should preserve G³ continuity (position, first, second, and third derivative matching at blade-to-shroud junctions) — G² is acceptable for the aerodynamic surfaces; G³ is required only at the structural joints.

Python Geometry Generator

File rotor_geometry.py (included in this repository) generates the full 3-blade surface point cloud from the contact-geometric equations above. Run with Python 3.9+ and NumPy: