System reliability analysis - Reliability computation of modular uninterruptible power supply (UPS) with decentralized parallel architecture (DPA) - ABB Power Protection

Oct 26, 2026

Upload by ahduag

White paper analyzing the system availability and reliability of ABB modular uninterruptible power supplies using a decentralized parallel architecture (DPA). The work defines the mean time before failure (MTBF) based on component-level MTBF/MTTR values across rectifiers, batteries, inverters, communication, protection elements, control boards, fans, and repairable parts. It applies reliability block diagram principles, considers N parallel modules and redundancy options (N-1/N, N+0, N+1) with common or separate battery banks, and presents calculated MTBF results for multiple configurations.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages5

Reading Time13 min

AccessFree to download

Introduction

Principle of the analysis

Reliability computation approach

Results and discussion

— WHITE PAPER System reliability analysis Reliability computation of modular uninterruptible power supply (UPS) with decentralized parallel architecture (DPA ™) ABB Power Protection Authors: Silvio Colombi, Consulting Engineer, Electrification, ABB Diana Garcia, UPS Product Manager, Smart Power, Electrification, ABB 1 Introduction ABB’s modular uninterruptible power supply (UPS) is based on ABB’s decentralized parallel architecture. With this architecture, each power module is equipped with its own rectifier, battery charger, inverter, static bypass and control logic for autonomous operation. Please refer to figure 1. Figure 1. Patented decentralized parallel architecture The patented DPA architecture maximizes power availability due to redundancy for critical components and fault tolerance. The autonomous operation of the power modules allows on-line, hot-swap ability for continuous uptime and automatic isolation in case of failure. A modular UPS system consists of parallel power modules with common or separate batteries and, potentially, internal redundant power capabilities. The reliability of the modular UPS system depends on the specific configuration selected as well as on the reliability of its components. In order to compute the system availability and reliability of an ABB modular DPA UPS, the mean time before failure (MTBF) of the system is determined, considering many sub-systems: rectifier, battery, battery converter inverter, power supplies, contactors, static switch module, redundant internal communication, control boards and fans. The MTBFsof the individual sub-systems is determined from field data related to 15 years of experience from an installed base of around 100,000 manufactured UPSs with similar design topologies and concepts. For brand-new designs, theoretical calculations are made to evaluate the reliability of the upgraded sub-system, starting from field data of similar consolidated designs and considering the changes. In fact, the Telcordia/Bellcore methodology (*), based on the component counts, does not reflect the internal circuit topology or the firmware aspect. (* ) The Telcordia document "Reliability Prediction Procedure for Electronic Equipment" (document Issue 1) can be ordered directly from Telcordia Customer Service in New Jersey; Phone: (800) 521-2673 or (732) 699-5800; the cost is about $1,000. Note: the Telcordia document, Issue 1, is an update of the Bellcore document TR-332, Issue 6. 2 Principle of the analysis For the reliability analysis, the following conditions are considered: 1. Number of power modules in parallel 2. Internal redundant capacity N+0 and N+1 3. Common and separate battery banks 4. Decentralized static bypass switch Please refer to table 1. Parallel module N N N N Redundancy N-1/N N-1/NN/NN/N Battery Common Separate Common Separate Static bypass Decentralized (DPA) Decentralized (DPA) Decentralized (DPA) Decentralized (DPA) Table 1. Modular UPS configurations Figure 1 shows the corresponding block diagrams. N represents the number of DPA power modules installed in the UPS system. The notation “N-1/N” means that a redundant module is used and, therefore, only (N-1) modules out of N are required to supply the critical load. On the other hand,“N/N” means no redundant module and, therefore, all the N modules used need to operate. Figure 1 Block diagrams for the two topologies considered for the modular UPS system: (a) separate batteries for each module, (b) common battery for all the modules. N represents the number of modules installed in the UPS system. For each of the block diagrams shown in Figure 1, the reliability block diagram (RBD) is established. An exhaustive analysis of the complete system has been done, considering all the failure paths as well as the interactions between hardware, software and protection. The basic principle of the RBD is that every sub-system that is not redundant and

FAQs

What decentralized parallel architecture (DPA) feature drives higher UPS availability in the analysis?

Each power module operates autonomously with its own rectifier, battery charger, inverter, static bypass, and control logic. This enables online hot-swap for continuous uptime and automatic isolation upon failure, improving availability through redundancy and fault tolerance.

Which input conditions are considered when modeling reliability for the modular UPS system?

The analysis considers the number of power modules in parallel, internal redundancy capacity (N+0 and N+1), whether battery banks are common or separate, and the use of a decentralized static bypass switch.

How are MTBF and MTTR used to compute system reliability in the presented method?

Component MTBF values are derived from field data and theoretical calculations for upgraded subsystems, while MTTR values are assigned to repairable parts. A reliability block diagram treats non-redundant, non-isolatable subsystems as series elements, then applies reliability equations to obtain configuration-level MTBF.

Oct 26, 2026

Upload by ahduag