Pre-Conference Workshop – Real Availability

The design, construction, operation, and maintenance of mission-critical facilities is a demanding art. 7x24 Exchange members know from first-hand experience that end-to-end reliability requires attention to all the details while never losing sight of the big picture. Most of the specialized knowledge and training is learned on the job – until now. Steve Fairfax and MIT Professor Mike Golay have distilled decades of applied reliability theory and practice into a one-day seminar – Real Availability. Developed especially for 7x24 Exchange International, Real Availability will both challenge and change the way you think about reliability and availability.

View Session Website

Steve A. Fairfax, President, MTechnology, Inc.
Michael Golay, Ph.D., Professor, MIT Nuclear Engineering Department

Tutorial Session A: Fundamentals of Data Center Commissioning

This session will explore the key elements and issues that a project team needs to address in order to deliver an effective data center commissioning program. It will cover language, objectives, methods, team member roles, and success criteria that characterize an outstanding program. The session will include real-life examples of findings and corrective action. Participants are invited to submit examples of facility issues they have experienced so that the session can address how a commissioning program might prevent similar occurrences in future projects.

Dave DiQuinzio, Principal, Strategic Facilities, Inc.

Tutorial Session B: Mission Critical Fire Protection Systems, Common Problems and Solutions

This presentation will look at common problems with fire protection systems experienced by the owners and users of mission critical facilities. False alarms, accidental clean agent releases and sprinkler system corrosion issues are included. This session will help managers take steps to determine if they are at risk and make adjustments before problems begin.

Brian K. Fabel, P.E., Director, National Accounts, Orr Protection Systems, Inc.

Tutorial Session C: American Express – Leveraging Dual Power for Availability SLA

This session will share how American Express’ Data Center Facilities have established variable availability SLAs for dual power complying and non-complying equipment installations and ownership. This discussion will include a 2005 outage case study demonstrating the major payoff of variable availability SLAs. John will discuss in detail the range of process, procedures and quality control measures that American Express’ Data Center Facilities have created to fully leverage dual power. Key points will also focus on benefits from American Express’ Global Standards and Approved Exception Process, CIO and Senior Leadership support principals, and investment trade-offs in single vs. dual powered technology buying decisions. Prime takeaways will include first steps for securing Senior Leadership support, dual power standards, plus how and where to start implementing variable SLAs within your data center services.

John Jackson, Director of Raised Floor Infrastructure Services, American Express

Tutorial Session D: Fundamentals of UPS Batteries in Data Centers

This tutorial will outline the evolution of the UPS battery choices available today and their main selection considerations. The linkage of maintenance requirements and failure modes will be explored, and a simple statistical assessment of battery reliability will be discussed. The session will conclude with a discussion of the possible impact of emerging battery technologies on the UPS market.

Jim McDowall, Business Development Manager, Saft America, Inc.

Tutorial Session E: Fluid Mechanics 101:
Understanding Cooling Airflow via Computational Fluid Dynamics

This tutorial session will begin with basic concepts of air velocity, airflow rate, pressure, and temperature distribution as applied to raised-floor data centers. You will be shown why the flow distribution through the perforated tiles is usually not uniform. It is governed by the air velocity and pressure variation under the raised floor. Computational Fluid Dynamics (CFD) can be used to calculate the complete velocity and pressure field and thus we can predict the airflow coming out of each perforated tile. CFD is a powerful technique to analyze a flow problem of any complexity. The CFD simulation will allows us to study the effect of variables such as: layout of the CRAC units and the perforated tiles, the height of the raised floor, and the presence of obstructions under the raised floor. Once the flow rates through the perforated tiles are determined, the next step is to apply CFD to the above-floor space to obtain the distribution of air velocity and temperature as the air moves through the server racks and back to the CRAC units. Many examples will be presented to develop an understanding of the physical processes and to draw practical conclusions. The tutorial will show how to create a computational model of a data center layout and calculate the corresponding airflow and temperature distribution.

Suhas V. Patankar, Ph.D., Professor of Mechanical Engineering, University of Minnesota and President, Innovative Research, Inc.

Keynote Address: American Express – Is there a difference between "7 x 24" and "always on"?

In this case study, John Ryan will review how translating system diagnostics into business metrics created real value. John is responsible for running a worldwide, always-on network for payments, travel and banking businesses. By understanding how to translate technical diagnostic information about the network into business relevant metrics, a richer understanding of priorities led to breakthroughs in investment strategies. John will share ideas on how to apply similar reviews to other "always on" infrastructure platforms.

John Ryan, Vice President, American Express Technologies

MIT – Defining and Base Lining Metrics Accurately for IT/Facilities

The high reliability electric power supply/IT world is greatly concerned with rules of thumb and metrics to monitor in seeking high performance. Looking to future outcomes we cannot know what will occur, but we can attempt to make the probability of mission success – the reliability – as high as possible, given available resources. Many factors contribute to high reliability, including low component failure frequency, avoidance of common cause failures, health monitoring and rapid intervention or repair, system architecture, and human and organizational performance. Each such factor has metrics indicating its condition (e.g., surveillance testing interval, degree of redundancy) that can be used to improve that factor’s contribution to good overall performance. However, no single metric is sufficient to indicate that an IT facility is being operated as well as possible, rather it is the integrated contribution of all such factors that determines overall performance outcomes. The important use of individual metrics is in a systems-based understanding of their combined contributions, where these can be balanced against the resources demanded to obtain them. It is such systems thinking that is essential for obtaining good performance, rather than a focus upon a small number of key performance metrics. Examples provided in previous 7x24 conferences (e.g., Fairfax and Golay, 2006) illustrate how such systems thinking can be implemented. Ref. S Fairfax and M Golay, "Optimal Test Strategies for Standby Diesel Engine-Generators" presented at the Spring 2006 7x24 Exchange National Conference.

Michael Golay, Ph.D., Professor, MIT Nuclear Engineering Department

Avaya – Squeezing the Last Nine in Reaching Five Nines Availability

Communications Manager is the 'brain' that controls the network of Avaya IP telephony gear which is deployed globally in more than 90% of Fortune 500 companies. To provide highly reliable service to end users there are systematic analysis and development done over many years to the Communications Manager system. These go beyond traditional end-to-end reliability measures and provide for call preserving upgrades, business continuity features and recovery from large numbers of failures. This presentation will describe the various initiatives underway over the years and the current challenges to reach the Five Nine availability for end users.

Dr. T.K. Srinivas, Director, Avaya Labs

SUN Microsystems – Analysis of Field Data for Repairable Systems

The analysis of field data on repairable systems often involves estimating the mean time between failures (MTBF). There are significant assumptions and potentially serious shortfalls surrounding common MTBF practices. Alternative, parametric modeling approaches are typically too complex to gain wide acceptance into the day-to-day reliability monitoring operations of a company. The purpose of the talk is to present simple graphical methods for analyzing and monitoring the reliability of repairable systems. This presentation will illustrate techniques based on non-parametric methods which have been successfully used within Sun Microsystems to transform the way the reliability of repairable systems is analyzed and communicated to management and customers.

David C. Trindade, Ph.D., Distinguished Engineer, Sun Microsystems, Inc.

Eaton Electrical – Harmonic Solutions for Mission Critical Applications

Problems associated with harmonic distortion are well understood for many power system applications. However, finding the right solution is challenging. There are many different technologies to choose from, each with specific technical and economic advantages. Data Centers and other Mission Critical applications require further considerations most of which pertain to the reliability of the solution and its effect on the overall power system reliability. This presentation will provide recommendations for reducing harmonic distortion, improving system capacity and improving system reliability while evaluating economic considerations using case study examples. Special considerations regarding the effects of harmonic loads on generators will be discussed. Finally, while reliability is of utmost importance at Mission Critical locations, optimizing power flow, saving energy and reducing the loading on HVAC systems can all result in savings (and potentially improved reliability). Some harmonic solution providers claim that their product will save energy as well as reduce harmonics. This paper will discuss these claims and the actual expected savings from these solutions.

Daniel J. Carnovale, P.E., Power Quality Solution Manager, Eaton
David G. Loucks, P.E., Power Solutions Manager, Eaton

Case Study: The Proof’s in the Power Bill:
Verified 10- 20% Energy Cost Reduction Using DC @ Data Centers

Data center UPSs convert utility AC to DC, then back to AC again for power supplies that convert it back to DC yet again. If you were to eliminate all those energy wasting, heat generating steps, your power bills would instantly drop 10-20%...and that's not even counting the cooling savings. With the active participation of Sun, Intel, IBM, Cisco and other industry giants, the Department of Energy’s Lawrence Berkeley Lab put together a side-by-side pair of blade test-beds at Sun Microsystems to prove the energy conservation and reliability enhancement of the DC distribution approach. Findings will be presented by:

Mike Bushue, Senior Staff Engineer, Sun Microsystems
William Tschudi, Environment Energy Technologies Project Manager, Lawrence Berkeley Laboratory
Brian Fortenbery, Engineering Manager, EPRI Solutions
Johnny Gonzales, Regional Sales Director, Pentadyne Power Systems

Tutorial Session A: American Express – Leveraging Dual Power for Availability SLA

This session will share how American Express’ Data Center Facilities have established variable availability SLAs for dual power complying and non-complying equipment installations and ownership. This discussion will include a 2005 outage case study demonstrating the major payoff of variable availability SLAs. John will discuss in detail the range of process, procedures and quality control measures that American Express’ Data Center Facilities have created to fully leverage dual power. Key points will also focus on benefits from American Express’ Global Standards and Approved Exception Process, CIO and Senior Leadership support principals, and investment trade-offs in single vs. dual powered technology buying decisions. Prime takeaways will include first steps for securing Senior Leadership support, dual power standards, plus how and where to start implementing variable SLAs within your data center services.

John Jackson, Director of Raised Floor Infrastructure Services, American Express

Tutorial Session B: Understanding Cooling Airflow via Computational Fluid Dynamics

This tutorial session will begin with basic concepts of air velocity, airflow rate, pressure, and temperature distribution as applied to raised-floor data centers. You will be shown why the flow distribution through the perforated tiles is usually not uniform. It is governed by the air velocity and pressure variation under the raised floor. Computational Fluid Dynamics (CFD) can be used to calculate the complete velocity and pressure field and thus we can predict the airflow coming out of each perforated tile. CFD is a powerful technique to analyze a flow problem of any complexity. The CFD simulation will allows us to study the effect of variables such as: layout of the CRAC units and the perforated tiles, the height of the raised floor, and the presence of obstructions under the raised floor. Once the flow rates through the perforated tiles are determined, the next step is to apply CFD to the above-floor space to obtain the distribution of air velocity and temperature as the air moves through the server racks and back to the CRAC units. Many examples will be presented to develop an understanding of the physical processes and to draw practical conclusions.

Suhas V. Patankar, Ph.D., Professor of Mechanical Engineering, University of Minnesota and President, Innovative Research, Inc.

Keynote Address:
Breaking Cycles: A Rags-to-Riches Story: From Homelessness to Wall Street

Chris Gardner is the head of his own brokerage firm and owner of three homes, a collection of tailored suits, designer shoes, and Miles Davis albums – but just 20 years ago, Gardner was homeless, carried all his possessions on his back, and lived – on occasion – in a bathroom at a Bay Area Rapid Transit station in Oakland, California. A true testament to the perseverance it takes to make it from "rags-to-riches", Gardner tells his story of overcoming obstacles, "breaking cycles", and the motivation it takes to make your own success. Listen as he recounts his fears of failure, exhaustion of pursuit, and great appreciation to those that helped him. Gardner's storytelling captures his audience completely, and his humor lights up the room.

Chris Gardner

Cognio – Why Best Effort is Not Good Enough

Wireless technology is everywhere and it's changing the way content and network providers create, package and distribute services, as well as the way consumers use those services. For businesses, wireless technology has created a fundamentally un-tethered world with Wi-Fi work zones that enable workers to better collaborate, interact and receive information. This wireless revolution has the potential to leapfrog the struggling wired world and take full advantage of innovative new technologies and applications that will lay the foundation for a clean and clear Wi-Fi infrastructure capable of delivering premium services. These services will not only radiate throughout the enterprise, but also the wireless Internet service provider markets, creating new opportunity for revenue creation and customer traction unavailable in the wired world, especially as the industry moves to embrace fixed/mobile converge as a means to keep customers connected. Currently most of these services must be delivered over the unlicensed "junk" bond making it difficult to deliver the requisite guarantees needed to support advanced services. As a result this revolution can only be realized when the performance, security and reliability of the Wi-Fi network is better than best-effort. Neil Diener will discuss the technical and business challenges of WiFi-based service delivery as well as discuss how Wi-Fi services play an integral role as the wired and wireless worlds begin to merge. Neil will also highlight the security concerns to be aware of in any wireless networks, such as back doors, and denial of service attacks. He will outline how enterprises can address critical WLAN problems dynamically and can also set policies that eliminate or control interfering devices that might be clogging the WLAN spectrum.

Neil Diener, Chief Technology Officer and Co-founder, Cognio

Breakout A: Watt Cost for High Density Cooling? A Cost Study of the Traditional Hot Aisle/Cold Aisle Heat Dissipation Methods Versus Active and Water Cooled Enclosure Solutions

Today’s standard method for managing heat loads in today’s MCDC’s (Mission Critical Data Centers) may no longer lend itself to the extreme heat loads produced in today’s high density compute environment. Enclosure cooling technology has come under intense scrutiny due to the extreme operating conditions of today’s data centers. This study will compare the build costs (CAPEX) and operating costs (OPEX) for enclosures populated from 3kW to 30kW utilizing (3) methods common to the data center. Hot Aisle/Cold Aisle (utilizing passive enclosures), Active cooling enclosures (utilizing cold aisle/work aisle) and Water Cooled enclosures. This study will present a clear and concise cost per kW for all (3) methods

John Consoli, VP Global Sales, AFCO Systems
James Fulton, Ph.D., Faculty Mathematics, Suffolk County Community College

Breakout B:
Fires in the Mission Critical Facility – A Look Back at Causes, Responses and Lessons Learned

This presentation will showcase fires experienced in Mission Critical Facilities in the previous 12 months. A number of cases will be looked at and include the nature of the fire, the response to it, and any lessons learned. In addition to recent fires, a number of infamous and/or interesting cases will also be presented.

Brian K. Fabel, P.E., Director, National Accounts, Orr Protection Systems, Inc.

The Caterpillar 6 Sigma Journey

This presentation will provide the audience a brief background of our company and explain why we needed to change our culture. The background behind the 6 Sigma methodology that was chosen to move us forward will be reviewed, along with the critical elements and metrics that we used to track our deployment and monitor those areas that needed modification or improvement. We will also cover how we are using 6 Sigma to change the way we work, our culture, and how we’re using it to drive execution of our company's Vision 2020.

Curt L. Wegener, 6 Sigma Division Champion, Caterpillar Inc.

Tutorial Session: Fundamentals of UPS Batteries in Data Centers

This tutorial will outline the evolution of the UPS battery choices available today and their main selection considerations. The linkage of maintenance requirements and failure modes will be explored, and a simple statistical assessment of battery reliability will be discussed. The session will conclude with a discussion of the possible impact of emerging battery technologies on the UPS market.

Jim McDowall, Business Development Manager, Saft America, Inc.

Keynote Address:
IBM – A Case Study in High Density Cooling: A Practical Application for Water in the Data Center

Most data center owners today cite challenges in space utilization, power delivery, cooling, and even structural loading in their facilities. The real life case study of a computing cluster at Georgia Tech demonstrates a cost-effective, forward-looking solution for two of the above parametric challenges: space utilization and cooling. A water based rack door heat exchanger was deployed as the centerpiece of a cooling solution for an area where the power/cooling density was 10 times greater than the capabilities of the rest of the facility. The solution was pivotal in creating the showcase-caliber facility that was desired by the end user.

Bret W. Lehman, P.E., Senior Engineer, IBM Corporation
Bartosz Ilkowski, Senior Research Technologist, Georgia Institute of Technology
Stephen L. Peet, Manager Technical Support, Bell South RDC Atlanta
Steven M. Battenfeld, P.E., Mechanical Department Head, Minick Engineering, Inc.

Case Study: USPS Processing and Distribution Center

After 70 years of continuous operation at a single location in downtown Philadelphia, the USPS was eager to modernize their aging mail distribution operation and gain easier access to major interstate arteries and air transportation. US Postal Service’s new Processing and Distribution Center will not only cut costs and improve service for the federal agency, but will work as a model to do so at USPS facilities all across the country. Ensuring the safe passage of 4,000 employees and 8,000,000 pieces of mail in and out of the facility every 24 hours presented some substantial challenges including security of personnel and process, operational reliability and maintenance in a low cost model environment.

Frank Nemia, P.E., Engineering Design Principal, Kling
Richard Farley, AIA, P.E., Principal, Project Director, Kling
James Binkley, FAIA, Senior Architect, U. S. Postal Service Headquarters

Electrical Failures: The Good News

This presentation consists of several case studies involving electrical power system equipment failure investigations. Each case study includes photos of the equipment and a discussion of the cause. Photos show failures involving transformers, fuses, breakers, switchgear, meters, busduct and other power system equipment. Lessons learned from these failures have been condensed into a handy checklist to improve electrical equipment maintenance and system reliability. This checklist is a valuable tool for those responsible, maintenance, operation or design of electrical facilities.

Ashley Harkness, Manager of Laboratory Services, Electrical Reliability Services