Product Requirements Document (PRD) and Technical Build Plan for EVChargeModeler Web App
- Overview
Product Name: EVChargeModeler Purpose: A web app for financial modeling of commercial EV truck charging stations (4-8 chargers, focused on California/Southwest) with project management features, including uploading and parsing contractor Statements of Work (SOWs) for expense comparison. Built for developers, fleet operators, and consultants. Target Users: Project managers, financial analysts, infrastructure developers; multi-user with roles (admin, editor, viewer). Value Proposition: Streamlines EV charging financial modeling (inspired by Atlas, NREL EVI-FAST, ANL CHECT) with CA-specific incentives (e.g., EnergIIZE, LCFS) and integrates SOW ingestion for cost alignment and project tracking. Development Environment: Documentation and initial code via Grok; development in Replit for collaborative prototyping and deployment. Scope: MVP includes financial modeling (DCF, NPV, IRR, LCOC) and basic SOW upload/parsing/comparison. Future phases: advanced analytics, API integrations. Out of Scope: Full CRM, hardware simulations, legal contract generation.
- Functional Requirements 2.1 Financial Modeling Inspired by Atlas (DCF, 100+ inputs), NREL EVI-FAST (MDHD risk analysis), and ANL CHECT (LCOC).
Model Creation: Wizard-style UI: Inputs for project scale (4-8 chargers), location (CA/Southwest with pre-filled incentives), charger types (150-350kW+ DCFC), utilization (40-60%), energy demand. Categories: CapEx ($100k-$250k/charger, installation, grid), OpEx (10-15%/year), Revenue ($0.40-0.60/kWh, subscriptions, LCFS $50k-$200k/year), Incentives (EnergIIZE up to 75%, PG&E/SCE rebates). Horizon: 10-15 years; discount rate 8-10%.
Calculation Engine: Methods: DCF for NPV/IRR/Payback; LCOC ($/kWh = Total Costs / Total kWh); sensitivity analysis (tornado charts). Key Formulas: NPV = Σ [Cash Flow_t / (1 + Discount Rate)^t] - Initial Investment IRR = Iterative solver (Newton-Raphson) LCFS = (Energy in MJ * EER * CI Factor) / 1,000,000 * Credit Value ($100-200)
Outputs: Dashboard with NPV, IRR, Payback, LCOC; charts (cash flow, sensitivities); exportable financial statements.
Customization: 100+ inputs (e.g., market growth 8%, equipment $6,000/kW USD); scenario comparison (side-by-side). Incentives: Auto-populate CA programs (NEVI $384M, CALeVIP, LCFS); stacking logic for 40-80% CapEx reduction.
2.2 Project Management
SOW Upload/Parsing: Formats: PDF, Excel, Word (drag-drop or file picker). Parsing: OCR (Tesseract.js) for PDFs; extract expenses, timelines, scopes; manual override for errors. Features: Tag to projects, version control, search by contractor/amount.
Comparison Tools: Layer SOW costs into models (e.g., override CapEx). Side-by-side: Model vs. Bid (highlight >20% variances). Alerts: Flag budget discrepancies.
Project Dashboard: List view: Projects (Planning, Bidding, Active). Collaboration: Share models/SOWs; comments. Export: PDFs/Excels of models, comparisons.
2.3 UI/UX
Dashboard: Project list, quick-start model button. Workflow: Tabs (Inputs, Calculations, Outputs). Visuals: Charts (Chart.js for cash flow, D3.js for sensitivities). Accessibility: Mobile-responsive, WCAG-compliant. Onboarding: Tooltips with CA incentive examples.
2.4 Non-Functional Requirements
Performance: <2s calculations for 100+ inputs; <500ms page loads. Security: OAuth/JWT auth; encrypted SOW uploads; role-based access. Scalability: 100+ users; Replit-hosted for MVP, scalable to AWS/GCP. Storage: Retain models/SOWs; export/backup options.
- Technical Build Plan (Optimized for Replit) 3.1 Tech Stack
Frontend: React.js (via CDN: cdn.jsdelivr.net/npm/react, react-dom) with JSX, Tailwind CSS (via CDN). Reason: Lightweight, Replit-friendly; Tailwind for rapid styling; Chart.js for charts. Backend: Node.js with Express (Replit’s default Node env). Reason: Handles file uploads, parsing; WebSockets for real-time collab. Database: SQLite (Replit’s built-in storage). Reason: Simple for MVP; swap to PostgreSQL/MongoDB for production. File Parsing: pdf.js for PDF extraction; xlsx for Excel; Tesseract.js for OCR. Reason: Browser-based, Replit-compatible. Calculations: math.js for DCF/NPV/IRR (e.g., Newton-Raphson for IRR). Reason: Lightweight, no heavy dependencies. Deployment: Replit’s built-in hosting (free for MVP). Reason: Instant deploy, team access. Other: Firebase Auth (via CDN) for user login; no
tags (Replit’s sandbox restrictions).3.2 Replit-Specific Optimizations
Lightweight Dependencies: Use CDNs (cdn.jsdelivr.net) for React, Tailwind, Chart.js to avoid Replit’s storage limits (500MB free tier). File Handling: Store SOWs in Replit’s /tmp or SQLite; limit uploads to 10MB to avoid slowdowns. Real-Time Collab: Leverage Replit’s multiplayer mode for team coding; use WebSockets (via socket.io) for live model sharing. Testing: Jest for frontend (pre-installed in Replit); manual testing for parsing due to Replit’s resource constraints. Limitations to Address: Replit’s CPU limits: Offload heavy calcs (e.g., Monte Carlo) to client-side math.js or post-MVP cloud (AWS Lambda). No persistent storage: Export models/SOWs as JSON for backup; use Replit’s repl.co URLs for sharing.
Migration Plan: Post-MVP, move to AWS/GCP for scaling (S3 for SOWs, EC2 for compute); Dockerize for consistency.
3.3 Core Code Snippets Below are starter snippets for key features, ready for Replit’s Node/React env. React Frontend (index.html)
<title>EVChargeModeler</title> <script src="https://cdn.jsdelivr.net/npm/react@18.2.0/umd/react.development.js"></script> <script src="https://cdn.jsdelivr.net/npm/react-dom@18.2.0/umd/react-dom.development.js"></script> <script src="https://cdn.jsdelivr.net/npm/@babel/standalone@7.20.0/babel.min.js"></script> <script src="https://cdn.tailwindcss.com"></script> <script src="https://cdn.jsdelivr.net/npm/mathjs@12.0.0/lib/browser/math.js"></script> <script src="https://cdn.jsdelivr.net/npm/chart.js@4.4.0/dist/chart.umd.min.js"></script> <script type="text/babel"> const { useState } = React;function App() {
const [inputs, setInputs] = useState({
chargers: 4,
capEx: 100000,
opEx: 0.1,
revenuePerKWh: 0.4,
utilization: 0.5,
horizon: 10,
discountRate: 0.08
});
const calculateNPV = () => {
let cashFlows = [];
for (let t = 0; t < inputs.horizon; t++) {
let revenue = inputs.chargers * inputs.utilization * 365 * 1000 * inputs.revenuePerKWh;
let costs = inputs.opEx * inputs.capEx;
cashFlows.push(revenue - costs);
}
return math.npv(cashFlows, inputs.discountRate) - inputs.capEx * inputs.chargers;
};
return (
<div className="p-4 max-w-4xl mx-auto">
<h1 className="text-2xl font-bold mb-4">EVChargeModeler</h1>
<div className="grid grid-cols-2 gap-4">
<div>
<label>Chargers (4-8):</label>
<input
type="number"
value={inputs.chargers}
onChange={(e) => setInputs({ ...inputs, chargers: Number(e.target.value) })}
className="border p-2 w-full"
/>
</div>
<div>
<label>CapEx ($/charger):</label>
<input
type="number"
value={inputs.capEx}
onChange={(e) => setInputs({ ...inputs, capEx: Number(e.target.value) })}
className="border p-2 w-full"
/>
</div>
</div>
<div className="mt-4">
<p>NPV: ${calculateNPV().toFixed(2)}</p>
</div>
</div>
);
}
ReactDOM.render(<App />, document.getElementById('root'));
Node Backend (server.js) const express = require('express'); const multer = require('multer'); const pdf = require('pdf-parse'); const app = express(); const upload = multer({ dest: '/tmp/uploads/' });
app.use(express.json());
app.post('/upload-sow', upload.single('file'), async (req, res) => { try { const data = await pdf(req.file.path); const expenses = data.text.match(/$[\d,]+/g) || []; // Basic regex for costs res.json({ expenses, raw: data.text }); } catch (error) { res.status(500).json({ error: 'Parsing failed' }); } });
app.listen(3000, () => console.log('Server on port 3000'));
3.4 Development Roadmap
Week 1-2: Setup Replit (Node/React); build frontend (wizard, dashboard); basic DCF calcs. Week 3: Backend for SOW uploads; SQLite for models. Week 4: Parsing (pdf.js, Tesseract); comparison UI; test in Replit. Post-MVP: Add incentives DB, sensitivities; migrate to AWS if needed.
3.5 Risks and Mitigations
Replit Limits: CPU/storage constraints; mitigate with client-side calcs, small file uploads (<10MB). Parsing Accuracy: OCR errors; add manual input UI. Team Collab: Use Replit’s multiplayer; sync via Git for production.
- Leveraging Existing Tools
Atlas/NREL/ANL: Publicly available user guides and PDFs provide input lists (e.g., Atlas’s 100+ variables), formulas (DCF, LCOC), and outputs (NPV, sensitivities). These are embedded above (e.g., wizard mimics Atlas tabs; LCFS formula from ANL). Workaround for Excel Access: Use published methodologies (e.g., ICCT cost breakdowns, NREL’s risk models) to replicate logic. If you obtain the files, upload them for me to parse (PDF/Excel supported). Data Sources: CEC 2024 ZEV Plan, NEVI CA Plan for CA-specific costs/incentives.
- Next Steps
Refine PRD: Add specific CA incentives (e.g., EnergIIZE rates) or SOW formats. Code Expansion: Request more snippets (e.g., SOW parser, sensitivity charts). Replit Setup: Create a Replit project; share invite for team collab.