Michele Blazek

Data center power requirements: measurements from Silicon Valley

Current estimates of data center power requirements are greatly overstated because they are based on criteria that incorporate oversized, redundant systems, and several safety factors. Furthermore, most estimates assume that data centers are filled to capacity. For the most part, these numbers are unsubstantiated. Although there are many estimates of the amount of electricity consumed by data centers, until this study, there were no publicly available measurements of power use. This paper examines some of the reasons why power requirements at data centers are overstated and adds actual measurements and the analysis of real-world data to the public policy debate over how much energy these facilities use.

Measurement of environmental impacts of telework adoption amidst change in complex organizations: AT&T survey methodology and results

Abstract Telecommuting practices and their environmental and organizational performance impacts have stimulated research across academic disciplines. Although telecommuting trends and impact projections are reported, few true longitudinal studies involving large organizations have been conducted. Published studies typically lack the research design elements to control a major confounding variable: rapid and widespread organizational change. Yet social science ‘Best Practices’ and market research industry quality control procedures exist that can help manage organizational change effects and other common sources of measurement error. In 1992, AT&T established a formal, corporate-wide telecommuting policy. A research and statistical modeling initiative was implemented to measure how flexible work arrangements reduce automotive emissions. Annual employee surveys were begun in 1994. As telecommuting benefits have been increasingly recognized within AT&T, the essential construct has been redefined as ‘telework.’ The surveys scope has expanded to address broader organization issues and provide guidance to multiple internal constituencies. This paper focuses upon the procedures used to reliably measure the adoption of telework practices and model their environmental impact, and contrasts those procedures with other, less reliable methodologies.

National and regional implications of internet data center growth in the US

The electricity consumption of data center hosting facilities (also known as server farms or server hotels) is a growing concern to utility demand forecasters, data center facility managers, energy analysts and policy makers. Combining estimates of US computer room floor space for hosting facilities with total computer room power density, we found that these US data centers in the aggregate required less than 500 MW of power in 2000, and used only about 0.12% of the electricity consumed nationwide in that year. In this paper, our order-of-magnitude estimate suggests that energy demands of these facilities do not represent an enormous new burden on the electricity industry as a whole. The fact that these facilities tend to be concentrated in certain areas, however, may mean that there will be significant regional electricity demands in some parts of the country. If combined heat and power (CHP) technologies were introduced to data center facilities on a large scale, initial calculations indicate that these facilities might even become net contributors of power to the electric grid.

Tale of two cities: environmental life cycle assessment for telecommunications systems: Stockholm, Sweden and Sacramento, CA

This paper summarizes the findings of a lifecycle impact assessment study consistent with the ISO 14041/ISO 14042 DIS standards. This study involved teams from key suppliers and telecom partners and describes the telecommunications systems in two cities: Stockholm, Sweden and Sacramento, CA, USA. The study is the first application of a system-life-cycle impact assessment (S-LCIA) in the telecom industry and uses the life-cycle stressor-effects assessment (LCSEA) methodology. A major goal of the study to provide a baseline impact profile of environmental impacts of the telecom system across the lifecycle. This S-LCIA baseline is it necessary precursor to evaluate the environmental implications of the use and growth of telecommunications and to develop possible environmental improvement strategies such as supply-chain management initiatives and applications of telecommuting and alternative work arrangements. The system wide assessment produced impact profiles for each city at three levels: system infrastructural operation, equipment manufacturing and full system including peripherals. Selected findings from this study are presented in this paper.

Life cycle management of personal computers in a service company

Over the past decade, many service companies have seen rapid advances in the use of personal computing equipment. Companies have upgraded their computing equipment in order to keep current with the technology and allow for the use of the newest software. With this change out of equipment, several issues arise including asset management, asset recovery, end-of-life management, and new procurement strategies. The focus of this paper is to describe AT&Ts efforts during 1997 regarding end-to-end life cycle management of personal computing equipment.

Integrating ISO 14000 Environmental Management Systems and design for environment at AT&T

This paper discusses what the ISO 14000 Series of Environmental Management Systems (EMS) Standards and the design for environment (DFE) movement are and how they complement each other in a holistic approach to environmental management. Both are based upon the premise of organized, systematic, and documented approaches which embrace, and are in part based, on the disciplines of total quality management (TQM). In addition, we review the drivers for and the specific programs and organizations within AT&T being established to meet AT&Ts goals for implementing EMS and DFE globally. This includes participation in the ISO 14000 standards development process, awareness training, facility plans, etc.

Standardization of the Life Cycle Environmental Performance in the Energy Sector: ASTM Draft Standard: E067110 Quantifying and Reporting the Environmental Performance of Electric Power Generation Facilities and Infrastructure; Implications to the Electronics Sector

Among the key suppliers to the electronics sector, the power generation sector has been standardizing its approach to reporting environmental performance across the broad portfolio. The draft ASTM Standard, E06710, on the quantification and reporting of environmental performance of the electric power generation facilities contains a technology-neutral, performance based approach to environmental performance reporting. With its emphasis on life cycle impact assessment methodology and comparisons to grid power, the standard provides a mechanism to address system wide improvement and to encourage improvement and efficiencies among all power generation technologies and end users. In addition, by conforming to Clause 9 of ISO 14042 LCIA standard for comparative assertions, the ASTM standard specifies an LCIA framework that requires impact classification/characterization on a spatial, temporal, and effect-intensity basis. This level of LCIA allows for an environmental accounting system with sufficient accuracy to create a quantitative environmental performance rating program. This paper illustrates comparison among various renewable and non-renewable power generation options. This EP-rating system will be useful tool for power customers and planners and will serve as the basis for transparent, accurate reporting of advantages/disadvantages of various generation options/portfolios. The results of these analyses will be useful to energy customers, policy makers, and emergent technology/distributed power generators. This approach can also be used to address engineering approaches to conformance to the restriction on hazardous substances as currently being standardized as part of ASTM F40.4

Predicting Future Power Requirements for the IT Sector: Making the Case for Including Life-Cycle Implications in Design of Servers

A recently published article in the Wall Street Journal cited increased power density requirements for servers on the order of 3000 W/sf. While existing datacenter power consumption patterns do not currently suggest such high power densities, emergent high performance servers may in fact require substantial changes to facility design of power and HVAC systems. The paper summarizes research to date on power requirements in datacenters and provide a roadmap for ongoing integrated research