Reserve Power System for Data Center Redundancy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data centers face high costs and complex maintenance procedures due to the need for full redundancy in primary power systems, which can be costly and require special procedures to switch between primary and backup power systems during failures.
Innovation Solution
A reserve power system that automatically switches computer systems from primary to reserve power independently, providing oversubscribed redundancy with automatic transfer switches and overload protection, allowing for efficient power distribution and reduced redundancy costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If full redundancy of primary power systems is provided, then reliability is improved, but capital costs and operational costs increase
Solution Approach 1:
The power distribution system is segmented into multiple zones with dedicated backup components for each zone. Instead of providing full redundancy across the entire system, each segment (zone) has its own localized backup PDU, reducing the total number of backup components while maintaining reliability within each segment.
Solution Approach 2:
Backup power capacity is allocated locally to specific zones based on their individual power requirements and criticality. Each zone receives backup capacity proportional to its needs rather than having uniform full redundancy throughout the system, optimizing resource utilization.
2Reliability
If full redundancy of primary power systems is provided, then reliability is improved, but device complexity increases
Solution Approach 1:
The complex backup system is divided into smaller, manageable zone-level backup PDUs. Each zone backup PDU is a simplified unit that handles only its local zone, reducing the overall system complexity compared to a monolithic full-redundancy system.
Solution Approach 2:
Automatic transfer switches are pre-configured with automated control logic that detects primary power failures and initiates transfer to backup power without requiring complex manual procedures. This preliminary automation simplifies the operational complexity of switching between power sources.
3Reliability
If full redundancy of primary power systems is provided, then reliability is improved, but ease of operation deteriorates due to special switching procedures
Solution Approach 1:
Automatic transfer switches are pre-programmed with control logic that automatically detects primary power failures and executes the switch to backup power. This eliminates the need for complex manual switching procedures, making the system easier to operate while maintaining reliability.
Solution Approach 2:
The backup power system is designed to self-activate automatically when primary power fails. The automatic transfer switches and control systems monitor power status and initiate backup power delivery without requiring operator intervention, simplifying operation.
4Quantity of substance
If oversubscribed reserve power system is implemented, then capital costs are reduced, but reliability may be compromised
Solution Approach 1:
Each zone receives backup power capacity tailored to its specific requirements rather than providing uniform backup capacity. This localized approach ensures that critical zones receive adequate backup capacity while less critical zones receive proportionally less, maintaining overall reliability while reducing total backup capacity needs.
Solution Approach 2:
The reserve power system is designed with oversubscription, where the total backup capacity across all zones is less than the total primary power capacity. This partial redundancy approach accepts that not all zones can be backed up simultaneously at full capacity, but ensures that individual zone failures are covered, reducing costs while maintaining acceptable reliability.
Data Source
AI summary
A system for performing computing operations in a data center includes one or more sets of computer systems, one or more primary power systems, and a reserve power system. The primary power systems include at least one power distribution unit that supplies power to at least one of the sets of computer systems. The reserve power system automatically supplies power to at least one of the sets of computer systems if a condition is met (such as a failure of the primary power system).