Battery-storage sizing and degradation analyser
A completed electrical planning study that sizes an LFP battery for a 150 kWp campus PV system while coupling hourly dispatch, prepared outages, expected degradation, and lifecycle cost.
Project definition
Problem statement
A battery selected from first-year savings can appear economical while capacity fade changes late-life reliability, peak reduction, operating cost, and replacement exposure.
The engineering problem is to compare power and energy ratings under the same hourly balance, outage, tariff, expected-life, and lifecycle-cost assumptions without treating a numerical result as a vendor quotation.
Project objectives
- Prepare an hourly campus load, PV, tariff, critical-load, and grid-availability profile.
- Evaluate 100 to 400 kWh and 50 to 150 kW candidates under explicit SOC, efficiency, power, and energy constraints.
- Couple each annual dispatch to the BLAST-Lite LFP expected-life model for fifteen years.
- Reject designs with unserved critical energy in any prepared outage year.
- Compare degradation-aware and nominal-capacity selection and test ageing-rate and temperature sensitivity.
Project structure
Project components
Profile preparation
Verifies the retained PVGIS response, converts local time, and creates the declared synthetic demand, tariff, and outage columns.
Sparse dispatch model
Solves hourly grid import, charge, discharge, stored energy, curtailment, unserved energy, and monthly demand peaks.
Expected-life model
Passes SOC, time, and temperature to the published BLAST-Lite 250 Ah prismatic LFP model.
Sizing and economics
Couples fifteen annual operating years and ranks only candidates that pass the prepared reliability rule.
Evidence pipeline
Retains hourly, annual, candidate, sensitivity, figure, test, Word, and PDF outputs.
Methodology
Project workflow
- 01Prepare the profile
Verify 8,760 PVGIS records and create the reproducible synthetic campus demand and outage cases.
- 02Solve each design
Run the same sparse hourly optimisation for all twelve energy and power combinations.
- 03Apply degradation
Use each SOC trace to update expected capacity and solve the next operating year.
- 04Filter reliability
Reject any candidate with unserved critical energy in one or more coupled years.
- 05Compare lifecycle cost
Select the least-cost reliable candidate and quantify the error from omitting degradation.
Demonstration scenario
A 150 kWp PV system supplies a declared synthetic campus demand under time-varying energy and monthly demand charges. Every 100 kWh candidate fails the prepared outage requirement. The 200 kWh, 50 kW design passes all fifteen years, reaches 83.27 percent expected state of health, and exposes an INR 376,647 cost error when degradation is ignored.
Engineering
Tools and method
- Tools
- The project uses Python, NumPy, Pandas, SciPy, BLAST-Lite, Rainflow, Matplotlib, Jupyter for subject analysis, simulation, and results.
- Electrical model
- Hourly power balance, SOC recursion, directional efficiency, grid availability, converter power, usable energy, and monthly peak constraints.
- Optimisation
- SciPy and HiGHS-compatible sparse linear programming with deterministic inputs and retained feasibility checks.
- Battery ageing
- BLAST-Lite expected cell capacity, annual feedback, 80 percent replacement rule, equivalent full cycles, and rainflow depth statistics.
- Lifecycle analysis
- Separate kW and kWh capex, fixed O&M, tariff escalation, discounting, replacement, and baseline comparison in Indian rupees.
- Verification
- Unit and branch tests, dependency consistency and audit, repository checks, exact container run, and full document render inspection.
Testing
Evaluation
Evaluation measures
- Hourly power-balance, state-of-charge, grid, power, and energy constraint satisfaction
- Maximum annual unserved critical energy across fifteen coupled years
- First-year and year-15 grid peak, equivalent full cycles, and operating cost
- Expected year-15 state of health and replacement year
- Degradation-aware lifecycle cost, NPV saving, and no-degradation omission error
- Sensitivity to fitted degradation rate and assumed cell temperature
Project boundaries
- Campus demand is synthetic and PV generation is PVGIS model output, not measured site data.
- Perfect annual foresight is a planning upper bound and the prepared outages are not a reliability forecast.
- BLAST-Lite predicts expected cell life and does not replace pack warranty, thermal, cell-balance, or failure evidence.
- The project does not control battery hardware or provide protection, interconnection, fire-safety, or construction design.
- Tariff and cost values are declared scenario assumptions and not a utility or vendor quotation.
Included
- 01150 kWp Hyderabad PVGIS source with retained provenance
- 02Declared synthetic campus demand and three prepared outage cases
- 03Twelve battery power and energy candidates
- 04Hourly dispatch with tariff, monthly demand, SOC, power, energy, and outage constraints
- 05Fifteen-year BLAST-Lite expected-degradation coupling
- 06Candidate, annual, hourly, sensitivity, and lifecycle result files
- 07Fifteen labelled figures and 52 annotated references
- 08Sixteen automated tests with 96 percent branch coverage
- 09Complete project files, models, calculations, and analysis material in a private GitHub repository
- 1083-page project documentation in PDF and editable Word formats
- 1117-page setup and usage guide in PDF and editable Word formats
Project record
No information is collected on this page.
- Permanent project ID
- GP-EE-0OFUKMH
- Catalogued
- 21 Aug 2026
- Completed
- 26 Aug 2026
- Verified
- 26 Aug 2026
- Demonstration
- Included in repository
Handover
After purchase
- 01Payment is confirmed
The project is marked unavailable and cannot be purchased again.
- 02Repository access is granted
The buyer's submitted GitHub account receives access to the private repository.
- 03The purchase record is delivered
The certification sheet is prepared from the reviewed buyer details and sent privately by email.