power back up 1-14052012
TRANSCRIPT
Newave SA, Switzerland – [email protected] - www.newaveups.com28/02/2006 ML3-49 1
UPS SOLUTION FOR HIGHEST UPS SOLUTION FOR HIGHEST
POWER PROTECTIONPOWER PROTECTION
AVAILABILITYAVAILABILITY
Newave SA, Switzerland – [email protected] - www.newaveups.com28/02/2006 ML3-49 2
PART ONEPART ONE
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 3
(n+1)- DPA . . .No Single Point of Failure. . .Highest Availability
Critical Application
““Mission critical applications”Mission critical applications”
requirerequire
high power availabilityhigh power availability
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 4
Why do we parallel UPS’s ? For Superior Reliability (Redundancy)
The answer to superior reliability requirements for business-critical applications is true redundancy, where all parallel modules equally share the load. If one module fails the remaining modules will continue to provide power protection.
For Power Upgrade (Scalability)To meet the power requirement demands of growing data-centers and telecomsystems it is possible to easily reconfigure an existing UPS-power system by simply adding one or more modules.
(n+1)- DPA . . . No Single Point of Failure . . . Highest Availabity
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 5
• Fictitious Parallel Architecture
- Hot-Stand-by (Cascade) Configuration
• Standard Parallel Architecture
- Parallel UPS
• Newave‘s Leading Concept
- Decentralized Parallel Architecture (DPA)
Evolution of Parallel Architecture
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 6
• Each UPS can support total load• No-load condition for No 2• Compromise solution - no parallel load
sharing• The overload capability is limited by the
capacity of one single system = 150%
typical • The system must execute 2 by-pass
operations under overload conditions in
series !
Hot-Stand-by Configuration Cascade system
Fictitious Availability
Mains
By-pass Line (1) Load
Mains
By-pass Line
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 7
100% load
50% load
50% load
• Load sharing with two units
• Each UPS can support total
load (if redundant)
• Full overload capability of
2 units = 300% typical
Standard Parallel Architecture
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 8
+
Decentralized Bypass
OR
Centralized Bypass
=
Two different parallel topologies
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 9
Parallel - UPS with Centralized Bypass
• Limited extension at high cost
• Low Reliability
• Extendible only with power interruption
• Unit splitting with high costs
• Additional cabling and cabinetSingle point of failure
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 10
• High reliability
• Easy to extend (scalable)
• Low cost for power extensions
• Easy downgrading from parallel to
single version is possible
Parallel - UPS with Decentralized Bypass
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 11
• Redundant by-pass
• Scalability
• No single point of failure
Parallel UPS + decentralized By-Pass offers:
What about control electronics?What about control electronics?
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 12
Single point of failure
• One defined UPS is leader
• Logic failure = System failure
• One critical UPS Element is not redundant
Parallel-UPS with Centralized Logic: “Master-Slave”
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 13
Single point of failure
• Only One Element can decide for all units
• Logic failure = System failure
• One critical UPS Element is not redundantC.B
Parallel-UPS with Centralized Logic: “Common Box”
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 14
Parallel systems offer increased reliability
They still have single failure points
It’s the intermediate step to reachIt’s the intermediate step to reach
redundancyredundancy
NOT EVERY PARALLEL UPS IS A REDUNDANT SYSTEM!
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 15
“Redundancy” means:The existence in the system of more than one possibility to
fulfil the demanded function - or - the system can afford at
least one single failure of any type
Redundancy degree is “n+x”:n = number of units / elements to fulfil the demanded function
x = number of units / elements which can fail without influence
n+1 = one unit / element can fail and the demanded function is maintained
“Redundant Parallel Systems”
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 16
Example:if the demanded function is fly, redundancy means:
• If one wing fails the demanded function is not maintained NO REDUNDANCY
• 3 engines can fails and the demended function is mantained REDUNDANCY N+3
Single point of failure
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 17
50% load
50% load
100 kVA
Demanded function is supply the load
100 kVA
Critical load = 100kVA
• Redundancy N+1
Redundancy in UPS means
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 18
Fault tolerance: A fault-tolerant or failure-tolerant system is a system which
continues its operation even if one or more parts of the system fail
Error masking: Capabilty to mask failures in order to avoid any influence on the
demanded function
Error detection: A sensor / supervisory-element to detect single failures
Error isolation: Capability to isolate faulty parts without disturbing the demanded
function
REDUNDANCY IN UPS MEANS
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 19
Fault Analysis - part identification: Capability to analyse a failure and identify the faulty part
Error removal: Replacement of faulty parts without disturbing the demanded
function
Return to normal mode: Normal operation will be restored without interrupting the
demanded function
REDUNDANCY IN UPS MEANS
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 20
!! WITHOUT INTERRUPTION !!
All these operations must be carried out withoutputting any strains on the load:
•Fault tolerance
•Error masking•Error detection
•Error isolation
•Fault analysis
•Error removal
•Return to normal mode
REDUNDANCY IN UPS MEANS
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 21
1. Decentralized Parallel Architecture (DPA)
2. Independent:
- Power Circuits (inverter,
rectifier, static bypass)
- Control Logic
- Regulation Logic
3. No single point of failure
4. Any UPS can be the logic leader (MultiMaster system)
5. No limitation in number of Parallel units
6. Perfect Load Sharing
Mains
Regulation
Logic
Control
Logic
UPS 1
Regulation
Logic
Control
Logic
UPS 2
Recifier Inverter
Recifier Inverter
Static Bypass
Parallel Regulation BusParallel Logic Bus
Critical load
Static Bypass
NEWAVE’S LEADING CONCEPT
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 22
Sequential softstart after utility recovery or gen-set start:
• no over-rating of gen-set• no disturbance of other connected loads
• no cables overheating
Connection to Wavemon Management Software allows:
• single unit access • single module access ( rectifier, inverter, battery)• remote monitoring and operation
DISTRIBUTED PARALLEL ARCHITECTURE DPA
Newave SA, Switzerland – [email protected] - www.newaveups.com28/02/2006 ML3-49 23
SECOND PARTSECOND PART
Newave SA – [email protected] - www.newaveups.com28/02/2006 ML3-49 24