Data Center Power Cost Estimation

NexGenIT's server room hosts 18 devices across 3 racks. The IT manager must estimate the monthly power bill using rated wattage, utilization factors, and the local electricity tariff — by hand, before checking against the actual utility meter reading.

# Aplly.xyz Case Study Submission

## Title
Data Center Power Cost Estimation

## Type
Technology/IT

## Difficulty
Intermediate

## Estimated Time
60 minutes

## Overview
NexGenIT's server room hosts 18 devices across 3 racks. The IT manager must estimate the monthly power bill using rated wattage, utilization factors, and the local electricity tariff — by hand, before checking against the actual utility meter reading.

## Case Details

Function Focus: Manual estimation, power-to-cost decomposition, utilization-adjusted scaling, PUE overhead

Scenario:
NexGenIT's finance team noticed the monthly electricity line item has crept up 22% year-over-year. The IT manager needs to produce a bottom-up estimate of what the server room should be costing, so finance can compare it against actual bills. They have an equipment inventory with rated power, estimated utilization per device type, and the facility's PUE (power usage effectiveness) ratio. No power monitoring software is installed yet — this must be calculated manually from nameplate specs.

Dataset Structure:
- Equipment inventory: 3 racks × 6 devices each, with Device ID, Role, Rated Wattage, and Estimated Utilization %
- PUE ratio for cooling and overhead
- Local electricity tariff (₹/kWh)
- Actual monthly bill (for comparison in the verification step)

Tasks:
1. For each device, compute the actual power draw (watts) = rated wattage × utilization %
2. Sum the actual draw across all 18 devices to get total IT load in watts
3. Convert total IT load to daily kWh, then apply the PUE multiplier to include cooling and overhead, then scale to 30 days
4. Multiply by the per-kWh tariff to get the estimated monthly cost — then compute the % difference vs the actual bill shown in the data
5. Only after submitting your manual estimate, re-derive the number using a spreadsheet or calculator and note any discrepancy

Expected Output:
A one-page estimation memo containing: per-device actual-draw table, total IT load subtotal, PUE-adjusted kWh total, estimated monthly cost, % variance from actual bill, and a brief explanation of the single largest source of estimation error.

Evaluation Criteria:
Correct utilization-adjusted power calculation per device, correct PUE application to total IT load (not per-device), correct unit conversions (watts → kWh), meaningful variance analysis against the actual bill.

## Data Sources

Equipment Inventory (Rack 1 — Web Tier):
| Device ID | Role | Rated Wattage | Utilization % |
|---|---|---|---|
| WEB-01 | Web Server | 450W | 65% |
| WEB-02 | Web Server | 450W | 65% |
| WEB-03 | Web Server | 450W | 55% |
| WEB-04 | Web Server | 450W | 55% |
| LB-01 | Load Balancer | 200W | 40% |
| LB-02 | Load Balancer | 200W | 40% |

Equipment Inventory (Rack 2 — Data Tier):
| Device ID | Role | Rated Wattage | Utilization % |
|---|---|---|---|
| DB-01 | Database Primary | 650W | 80% |
| DB-02 | Database Replica | 650W | 60% |
| STO-01 | Storage Array | 400W | 70% |
| STO-02 | Storage Array | 400W | 70% |
| BACKUP-01 | Backup Server | 350W | 30% |
| BACKUP-02 | Tape Library | 200W | 10% |

Equipment Inventory (Rack 3 — Application Tier):
| Device ID | Role | Rated Wattage | Utilization % |
|---|---|---|---|
| APP-01 | App Server | 400W | 60% |
| APP-02 | App Server | 400W | 60% |
| APP-03 | App Server | 400W | 45% |
| APP-04 | App Server | 400W | 45% |
| SW-01 | Network Switch | 150W | 80% |
| SW-02 | Network Switch | 150W | 80% |

Constants:
- PUE (Power Usage Effectiveness): 1.7
- Electricity tariff: ₹8.50 per kWh
- Days in month: 30
- Hours per day: 24
- Actual bill for this month: ₹62,400

## Solution Frameworks
Fermi-style decomposition, utilization-weighted estimation, PUE-adjusted energy costing

## Solver Guidance & Tutorials
Link to: "Estimating IT Infrastructure Operating Costs" tutorial

## What You'll Learn
- Bottom-up power cost estimation without monitoring tools
- The impact of utilization assumptions on total cost
- How PUE inflates apparent IT energy costs
- Reconciling bottom-up estimates against actual bills

## Tags
data center, power cost, energy estimation, IT operations, PUE

## Registration Links
- Register as Solver
- Register as Evaluator

Equipment Inventory (Rack 1 — Web Tier):
| Device ID | Role | Rated Wattage | Utilization % |
|---|---|---|---|
| WEB-01 | Web Server | 450W | 65% |
| WEB-02 | Web Server | 450W | 65% |
| WEB-03 | Web Server | 450W | 55% |
| WEB-04 | Web Server | 450W | 55% |
| LB-01 | Load Balancer | 200W | 40% |
| LB-02 | Load Balancer | 200W | 40% |

Equipment Inventory (Rack 2 — Data Tier):
| Device ID | Role | Rated Wattage | Utilization % |
|---|---|---|---|
| DB-01 | Database Primary | 650W | 80% |
| DB-02 | Database Replica | 650W | 60% |
| STO-01 | Storage Array | 400W | 70% |
| STO-02 | Storage Array | 400W | 70% |
| BACKUP-01 | Backup Server | 350W | 30% |
| BACKUP-02 | Tape Library | 200W | 10% |

Equipment Inventory (Rack 3 — Application Tier):
| Device ID | Role | Rated Wattage | Utilization % |
|---|---|---|---|
| APP-01 | App Server | 400W | 60% |
| APP-02 | App Server | 400W | 60% |
| APP-03 | App Server | 400W | 45% |
| APP-04 | App Server | 400W | 45% |
| SW-01 | Network Switch | 150W | 80% |
| SW-02 | Network Switch | 150W | 80% |

Constants:
- PUE (Power Usage Effectiveness): 1.7
- Electricity tariff: ₹8.50 per kWh
- Days in month: 30
- Hours per day: 24
- Actual bill for this month: ₹62,400

Fermi-style decomposition, utilization-weighted estimation, PUE-adjusted energy costing