KNGenuity, LLC
Engineering · Data Science

Engineering the
water-energy nexus,
grounded in data.

KNGenuity is a hydropower water-quality and data-science consultancy. Most water-resources problems aren’t bottlenecked by data or by engineering — they’re bottlenecked by the gap between them. I close that gap.

What I do → Start a conversation
Services

Three disciplines, one practice.

Most water-resources problems aren't bottlenecked by data or by engineering — they're bottlenecked by the gap between them. KNGenuity bridges that gap with doctoral-level depth on both sides, and teaches other engineers to close it themselves. Hydropower water quality is the primary domain, but environmental monitoring, resource optimization, and other data-intensive engineering challenges are in scope when the technical problem is real.

Hydropower & Water-Quality Engineering

Hands-on expertise at the intersection of generation, regulation, and the receiving stream — data-driven and built across years of direct practice on operating hydropower projects.

  • Dissolved oxygen (DO) assessment & enhancement strategy
  • Conceptual & preliminary design of aeration systems
  • Minimum flow system design & performance assessment
  • Thermal stratification, destratification, and downstream temperature impact studies
  • Tailrace water-quality monitoring & instrumentation review
  • FERC licensing & relicensing technical support
  • §401 Water Quality Certification support

Data Science & Analytics

Statistical rigor and modern machine learning, applied with engineering judgment — never as a black box. Every model is built to hold up to a regulator, an auditor, or a skeptical colleague. The methods are rooted in hydropower and water resources, but transfer readily wherever rigorous quantitative analysis is what's needed.

  • Statistical analysis of environmental & operational data
  • Time-series modeling for reservoir & river systems
  • Predictive DO forecasting & real-time operational decision support
  • Network and operational optimization using heuristic methods
  • Custom Python pipelines for monitoring & reporting
  • Publication-quality visualization for technical reports

Professional Training & Seminars

Half-day, full-day, and two-day seminars on AI-assisted engineering workflows for water resources professionals — built around a verification-first discipline, not a tool demo. Attendees leave with a repeatable process, not just a list of prompts.

  • AI-assisted data QA/QC and statistical-analysis workflows
  • Publication-quality visualization techniques for technical reports
  • Verification discipline for AI-generated engineering work
  • Hydrologic & weather-data analysis labs (USGS, NOAA/NWS)
  • Organizational AI-use policy development
  • In-house and open-enrollment sessions available on request
Approach

Decisions from data —
not opinions about data.

“Data without engineering judgment is noise. Engineering without data is guesswork. Both, or neither.”

Every KNGenuity engagement starts with the same question: what decision are we actually trying to make, and what evidence would change it? That keeps studies focused, models honest, and reports short.

The deliverable is not a dashboard or a deck. The deliverable is an answer you can act on.

Selected work

Methods that travel.

Hydropower is the home base, and the underlying analytical toolkit — statistical inference, machine learning, physics-based simulation, and heuristic optimization — travels well beyond it. Selected client engagements and active research.

Current Machine learning

Forecasting dissolved oxygen at hydropower plant discharges

Developing a machine-learning model to predict dissolved oxygen at hydropower plant discharges across multiple time horizons — letting operators think in terms of the upcoming DO trajectory, not just the current snapshot.

  • Time-series ML
  • 2-hour & 24-hour horizons
  • DO prediction
  • Hydropower operations
Current Measurement uncertainty

Building an uncertainty budget for a field flow measurement

Turbine discharge can’t be metered directly — it’s inferred from how much a known tracer mass is diluted downstream, which makes the result only as trustworthy as its uncertainty budget. I’m quantifying that budget for a client: propagating error through the measurement chain and separating systematic from random contributions across mixing, tracer behavior, and instrument calibration. The output is a ranked list of where the uncertainty actually lives, so effort goes to the terms that move the number rather than the ones that are easiest to fix. The same approach works for any measurement that has to hold up to a regulator or an auditor.

  • Uncertainty quantification
  • Error propagation
  • Tracer dilution
  • Instrument calibration
2003 Heuristic optimization

Optimizing a water-quality sampling network in Great Smoky Mountains National Park

Applied simulated annealing — a stochastic global-optimization heuristic — to redesign a water-quality sampling network for the park, balancing spatial coverage, redundancy, and access cost. The same class of methods (simulated annealing, genetic algorithms, tabu search) routinely solves siting, scheduling, and sensor-placement problems where exact methods don't scale.

  • Simulated annealing
  • Sensor placement
  • Spatial sampling design
  • National park hydrology
Research Digital twin

Digital twin of a single-unit hydropower facility

Building a physics-informed digital twin that integrates real-time SCADA telemetry and water-quality sensor data with an operational simulation model. The twin mirrors actual plant behavior — unit commitment, tailrace hydraulics, and dissolved oxygen dynamics — so operators can run what-if dispatch scenarios before committing to a generation schedule. Built initially for a single-unit facility; designed for broader application across multi-unit projects.

  • Digital twin
  • SCADA integration
  • DO dynamics
  • Operational simulation
  • Hydropower dispatch
In Development Thermal modeling

Longitudinal stream temperature model for a hydropower tailrace

Developing a reach-scale stream temperature model for a hydropower tailrace and the receiving river downstream. The model combines meteorological forcing, solar radiation loading, and discharge boundary conditions to simulate longitudinal temperature profiles at regulatory compliance points. Designed to support permit negotiations and to quantify thermal impact under varying generation and minimum-flow scenarios.

  • Stream temperature
  • Energy balance
  • Longitudinal profile
  • Thermal compliance
  • Generation scenarios
Live River monitoring

Live gage dashboard for the Black Warrior–Tombigbee–Mobile river system

A self-contained dashboard that pulls real-time stage and flow readings directly from USGS gages across the Black Warrior, Tombigbee, and Mobile river system. It runs entirely client-side in the browser — no backend, no database — refreshing automatically against USGS's public water-data API.

  • USGS real-time data
  • River gage monitoring
  • Client-side dashboard
  • No backend required

Opens in a new tab and loads live gage data on open.

Open the live dashboard →
About

A small practice,
by design.

KNGenuity is a single-principal consultancy. That means clients work directly with the engineer doing the analysis — not a project manager translating between them and a junior staffer. It also means the workload is deliberately limited. I take on engagements where the technical problem is substantive and the client is serious.

Based in the United States. Available for short technical reviews, multi-month studies, and ongoing analytics support.

Contact

Have a technical problem to solve?

Tell me briefly what you're working on. I respond to serious inquiries within two business days.