Affordable Redundancy In Low And Medium Power UPS Ratings  Redundancy has always been used for critical applications. Now, due to a technology breakthrough, the same high reliability that traditionally was only available in large UPS installations is affordable for small and medium UPS ratings. Whether you prefer a true On-Line or Line-Interactive UPS, redundancy can significantly lower the probability of a load loss. 

Introduction To MTBF 
  
The common acronym used to specify reliability is MTBF, which stands for Mean Time Between Failures. Don't bother asking a UPS manufacturer for the MTBF of a low power UPS because they usually don't publish it. Why? MTBF can be used as a weapon by the competition and it is extremely difficult to defend the calculations. Richard Melton in a paper in the Power Quality Solutions - September 1995 Proceedings entitled How Much Uptime Does Your System Require? says "Since there are no MTBF police, reliable, comparable MTBF numbers do not exist." Even without "reliable, comparable MTBF numbers," it is easy to show the benefits of redundancy for UPS products in the small to medium kVA range. Although the justifications for redundancy given for large UPS installations are in general applicable, special consideration must be given to clearly understand the effects for small UPS applications. 

What Does MTBF Really Mean? 

MTBF is the mean (average) time between failures. It does not mean that if the MTBF is one year (8760 hours) that the UPS is guaranteed not to fail for 1 year. Unfortunately the story isn't nearly so good. It means that there is a 63% chance that a specific UPS will fail in a one year period or that 63% of all the UPS installed will fail within a one year period. The probability of failure (Pf) is determined from the MTBF number using the well known formula: 

 
Pf = 1 - e-(t/MTBF)

Table 1 shows the probability of failure in one year and in three years for several MTBF values. As a point of reference, an MTBF of about 100,000 hours is achievable for a 1kVA UPS. 
  

Table 1. Probability of Failure
MTBF Pf in One Year Pf in Three Years 
8,760 hours (1 year)  63% 95%
25,000 hours (2.9 years)  30% 65%
50,000 hours (5.7 years)  16% 41%
100,000 hours (11.4 years) 8.4% 23%
200,000 hours (22.8 years)  4.3% 12%
400,000 hours (45.7 years) 2.2% 6.4%
 
High Mission Reliability 
  
The main reason for buying a UPS is battery backup. If a Line-Interactive UPS has failed, or an On-Line UPS is on bypass, battery backup cannot be provided. In simple terms, the UPS cannot do its job, cannot perform its critical mission. One way to quantify reliability is to estimate the likelihood that a UPS will be able to do its job and this is called Mission Reliability. The Mission Reliability of two UPS products can be compared using Mission MTBF numbers. A redundant UPS has a much higher Mission Reliability than a conventional UPS does. The following discussion explains why. 

Mission MTBF is determined by the following daunting formula: 
 

lmission =(s+1)Al/a where: 
    l = 1/MTBF of each module 
    A = n!as+1/(n-s-1)!(s+1)! 
    a = l (1-e-(l+m )t/(l+m ) 
    m = 1/MTTR (Mean Time To Repair) 
    t = system life time 
    n = total number of modules 
    s = Number of modules on Standby 
  
The mission MTBF of a non-redundant UPS is the same as the MTBF calculated using MIL-STD-217 or by another method. The mission MTBF of a redundant UPS is much higher because one module can fail and the UPS will still function. Two or more modules have to fail before a "mission" failure occurs.
 
The Importance Of The Mathematical (MTBF) Model Used 
  
In order to calculate Mission MTBF, the mathematical model used is very important. For example, if we determine that the non-redundant 6kVA UPS has an MTBF of 45,000 hours, the Mission MTBF is also 45,000 hours. If we assume that a 5kVA redundant UPS uses six 1kVA modules with a 4,000 hour MTBF each, the Mission MTBF is 5.3 million hours (600 years)! We know this is not realistic even though a very low MTBF was chosen for the modules! 

A more realistic model (See Figure 1.) is to assume that part of the UPS is simple and therefore very reliable (e.g. the Input/Output terminals, wire, etc.) and part is more complex and therefore more prone to failure. For the purpose of explanation, we can split the MTBF model for the conventional UPS into a simple section with 500,000 hours MTBF and a complex section with 50,000 hours MTBF. The MTBF when combining 500,000 and 50,000 hours is 45,000 hours. 
 

Figure 1. Non-redundant UPS Model
6kVA UPS, Load = 4.8kVA, Est. MTBF = 45K hours
 
The model for the redundant system is similar (See Figure 2.), but the complex parts now are redundant. To achieve redundancy, six 1kVA UPS modules are paralleled and a simple cabinet makes up the high-reliability part of the total UPS. In this model, the MTBF of the simple part of the UPS is estimated to be the same 500K hours and each module has an estimated MTBF of 100K hours. Because one module is held in reserve, to create a redundant system, the Mission MTBF is 494,000 hours. 
 
Figure 2. Redundant MTBF Model
 6kVA UPS, Load = 4.8kVA, Est MTBF = 494K hours

Note: For the sake of comparison, the MTBF of the AC input/bypass power source was not included. Eight hours was used as the MTTR input to the calculation for the redundant UPS. 

Why is the mission MTBF of the redundant UPS so much better. In simple terms, in a fault tolerant, modular, redundant UPS, the mission MTBF approaches the MTBF of the simple parts of the UPS, the wires and terminals, etc. This is because the failure of a single 1kVA UPS module will not prevent the UPS from continuing its mission. 

Benefits Of A Modular, Redundant UPS 

  • The technology break-through that makes redundancy affordable comes with the additional benefit of easy, fast, safe module swapping. The modules can be removed and installed while the UPS is supporting the critical load (Hot-swap capability). This means that the bypass (i.e. putting the load at risk) is not used to change a module, repair the UPS or do maintenance.
  • Because of the unique hot-swap capability and modular design, the removal and replacement of a module takes only a few minutes. (Assuming a spare module is available and close by.) Modules are light weight and can easily be replaced, returning the UPS to full power. If a replacement module has to be flown in, the UPS is still protecting the critical load. If a conventional UPS fails, the load is at risk if the power fails while waiting for the spare part.
  • If a module fails, the UPS continues its mission, which is to provide uninterrupted power. This is possible because PowerKinetics' PowerReactor series of UPSs are designed to be fault-tolerant. The state of the art, fault-tolerant, techniques used mean that the UPS will identify and isolate a module problem so quickly that the UPS output will not be disturbed.
Conclusion 
  
PowerKinetics is the true leader for high mission reliability in a UPS. No other UPS can match the price and performance of a PowerKinetics fault-tolerant, hot-swap capable UPS. Both true On-Line and Line-Interactive configuration are available, allowing the customer to choose his or her preferred design. Since the modules are common across kVA ratings and across single-phase and three-phase models, the end-user doesn't have to worry about stocking separate spares for each type and rating of UPS. With so many advantages to the user and world class R&D and customer support, PowerKinetics products are poised to change the UPS market.
 
Back to Technical Bulletin