Your green technology strategy should extend beyond screen savers and virtualization. Meet your growing information technology requirements while reducing the burden placed on our power grid.

The first two articles in the green data warehouse series focused on servers and the data center – providing a measurement framework and giving you tools for reducing your energy consumption. This article will provide a full end-to-end view of energy consumption, from the applications that drive your business to the power generation facilities supplying that energy.

As reported in my article The Green Data Warehouse, Part 1, IT is driving significant energy usage, estimated at 1.5 percent of all electricity consumption in the United States today. Current peak load consumption is estimated at 7 gigawatts (GW) requiring 15 baseload power plants. If current trends continue, this demand would rise to 12 GW by 2011, requiring an additional 10 power plants. A 2007 EPA report estimates that data centers in the United States have the potential to save at least $4 billion in annual electricity costs (a 25 percent reduction in energy consumption), with 80 percent or more savings possible when implementing “state-of-the-art” practices. Although a number of the “state-of-the-art” practices require major renovation of the data center, a more modest goal of 30-40 percent is still achievable without picking up a shovel. The EPA study identified “server consolidation” as a key factor in achieving these levels of reduction, but the study does not describe how that would be achieved.

While the EPA study takes an IT infrastructure-centric view to rising demand and associated reduction strategies, there are other drivers relevant to the discussion, namely:

• Business needs – data and transaction volumes are increasing due to competitive and regulatory pressures
  
  
• Regulation – a “pollution tax” in the form of CO₂ emission taxes or cap and trade agreements
  
  
• Politics – the current administration is backing a “smart grid” policy that has both cost-saving and revenue-generation opportunities

A holistic view stretching from the IT application to the power source will provide insight into the true source of power consumption from a business perspective and help identify specific best practices for aligning demand with supply.

From IT Application to Power Generation

An ecosystem is a group of living organisms functioning with the non-living elements in a given area. The green technology ecosystem aligns your IT needs with power generation, creating a harmony between demand and supply. In the IT world, applications such as data warehouses, CRM and ERP create demand for transaction processing and data storage – demands satisfied by servers usually housed in a specialized building known as a data center. The data center is constructed to optimize the housing of servers and other hardware components. The lifeblood of the data center is the local energy grid, 50 percent of which is generated by coal-fired power plants. Currently, most power is uni-directional, meaning power is supplied with little knowledge of the power consumption demands.

Figure 1 depicts the ecosystem and forms the basis for the discussion around optimizing the flow of energy through this system.

Figure 1 

The system is divided into four focus areas: Application Tuning and Rationalization, Data Compression and Transaction Acceleration, Resource Optimization, and Smart Grid Alignment. For each focus area, I will discuss the objective as it relates to energy efficiency, identify several best practices for achieving those objectives, and then recap the associated benefits.

1. Application Tuning and Rationalization

Best practices include:

The benefit is a reduction in transaction volumes and data storage requirements. Unlike data compression and transaction acceleration techniques detailed subsequently, this is achieved by eliminating the requirement to store a data element or process a transaction.

2. Data Compression and Transaction Acceleration

The objective is to store data in the most efficient manner, and to process transactions with the least number of machine cycles. In some cases, it may be cost prohibitive to rationalize the business requirements or optimize the application design. Either way, significant gains in footprint reduction can be achieved in this area.

Best practices include:

As with application tuning and rationalization, the primary benefit is a reduction in transaction volumes and data storage requirements. The difference is that this is achieved via optimization of data sets and transactions as opposed to eliminating them outright.

3. Resource Optimization

The objective is to optimize usage of available hardware resources, including servers, network equipment, disk drives and backup devices. The key here is breaking the one-to-one relationship between application and server, enabling capacity from multiple servers to be pooled and shared by many applications.

Best practices include:

The benefit is a reduction in new hardware purchases since existing resources are leveraged more efficiently. This also aligns with the EPA goal of server consolidation by leveraging existing infrastructure to service multiple applications.

4. Smart Grid Alignment

The objective is to align energy consumption requirements with power plant generation capabilities. Our nation’s power grid is mostly one directional in that power flows to consumers, but little if any information is returned to the source. In some cases, power consumers have become power generators, selling excess capacity back to the grid, but this is a small percentage of our generation capability. The move to a “smart grid” is on the current administration’s agenda, with funding earmarked within the new stimulus package.

Best practices include:

The benefits are lower kilowatt per hour (kWh) costs, and potential revenue generation opportunities by selling excess capacity from backup generators and thermo-capture devices.

Looking at any one of these areas in a vacuum can produce incremental improvements, but I believe exponential gains will only be achieved by analyzing the system in totality.

Business Case

The business case includes three elements:

Cost savings achieved via the following:

• Reduced energy consumption due to server efficiencies and data center resource optimization – techniques 1 & 2 identified ways to reduce the demand for resources by rationalizing business requirements and optimizing design, in addition to compressing data and accelerating transaction processing for the information that needs to be stored.
• 
  
• Smaller footprint reduces data center expansion pressure – the first three focus areas will result in a reduction in footprint that directly leads to reduced demand for more physical space.
  
  
• Reduction in new hardware and software purchases – the first three also contribute to a scale down in required hardware and in most cases associated software such as database and other packages priced on a per server or CPU basis.

Revenue generation achieved via the following:

• Adoption of “smart grid” capability could position you to sell excess capacity – data centers have several ways to generate energy, from old school generators to new age thermal-capture devices that turn heat generated within the data center into electricity.
  
  
• Federal tax credits – cap and trade laws and tax credits could be granted to incent companies to integrate with a smart grid.

Publicity and goodwill generated because of:

• “Green” initiatives that raise awareness in the public eye – consumers are becoming more aware of the efforts made by companies to enact “green” initiatives. Brand loyalty will be defined in part by your ability to articulate the steps you are taking to reduce your energy consumption.

Road Map

The following is a straw-man approach for launching a green technology ecosystem initiative. The focus is on quickly identifying one or more focus areas that will yield the biggest return in ongoing savings in relationship to the investment cost.

Assessment (2-3 weeks):

• Sample current hardware infrastructure and applications
  
  
• Assessment of future infrastructure growth plans
  
  
• Benchmark against industry standards
  
  
• Business case creation / Go/no-go decision

Implementation (2-3 months)

• Prioritize business case items
  
  
• Develop action plan and implement
  
  
• Test and monitor results
  
  
• Deploy ongoing measurement capability

Maintenance and tuning (ongoing)

• Periodic benchmarks against industry standards
  
  
• Adjustments to measurement metrics

Conclusion

The green technology ecosystem is a framework for assessing your energy consumption from business applications through power source. Taking an end-to-end view will focus your efforts on the areas likely to return the biggest reductions in energy consumption. This will also allow you to align your business objectives with the community goals of reducing the need for more power generation capacity, and may raise your profile in the eye of the consumer.

I did not cover the role that technology, and in particular data warehousing, plays in supporting this ecosystem, particularly the “smart grid.” My next article will discuss how technology can be leveraged to optimize the flow of energy from power generator to end consumer.

Recent articles by Rick Abbott

• Green Technology Infrastructure
• The Green Data Warehouse, Part 2
• The Green Data Warehouse, Part 1