Although not in the spotlight like the automobile and aviation industries, banks and financial services firms have sustainability issues firmly on their agenda.

Rising energy prices and the need to satisfy firm-wide sustainability objectives mean a need to minimise energy use through more efficient use and design of IT infrastructure and systems. And this is not limited to what a firm uses in house: best practice is demanded right down the supply chain.

Happily, this trend dovetails with a period in which banks and financial services firms are already looking to maximise IT by making best use of existing hardware and infrastructure. This is combined with a period of huge technological change where commodity systems, be that via a data centre or cloud, can offer huge cost, and by extension, energy savings.

But that is not to say the sustainability is merely a by-product of cost efficiency. In recent years legislation such as the UK Carbon Reduction Commitment (CRC), as well as general awareness, means that sustainability is now a desirable outcome in its own right.

Pankaj Kulkarni, an associate vice president at Infosys says: "Banks now know they can derive sustainability and cost efficiencies simultaneously, and that both are desirable. The banking industry is also actively looking to its partners to provide sustainable IT ecosystems. Although the conversation may start off on commercial terms, sustainability issues are a key differentiator - especially if they can be measured accurately."

On the agenda, then, is the drive to use IT sparingly; using, saving and storing only what is required and doing so in the most efficient way possible. Harnessing new technologies that allow for this while not upping either cost or energy use are a key part of the proposition.
Data management and storage is one key front end facet of reducing energy consumption. Good data management means less is storage required, thus using less energy.

Sarah Greasley, technical director at IBM financial services, says: "The focus is that if you can reduce and control the number of instances where the same data is saved then the requirement for storage also reduces." She cites a common example of an email going round with an attachment, which everyone downloads to their machine and then makes small changes to. "Instead, the document can be held on a web-based application to which everyone has access and can work on so that there is only ever one copy," she says.

Data storage is also all about how quickly archived data needs to be accessible and where it is best stored. Know Your Customer documentation, for example, or previous bank statements, are two obvious examples of data obviously needs that to be stored but does not need to be instantly accessible.

Virtualisation

Another key to enabling better management and storage is virtualisation. This is similar in principle to the cloud in that data and software are both held on central virtual servers, as opposed to on a desktop. Unlike a public cloud it does not involve data security or privacy issues because the virtualised service can still be hosted within a dedicated data centre. It is thus ideal for the financial services industry.

Ian Gausden, chief operating officer at VocaLink, says that the end user working on a virtualised machine would barely notice the difference. "The way it works is to install Secure Instant Messenger client units instead of PCs. The screen is a normal windows screen but everything is stored and saved virtually at a huge data centre or within a private or public cloud. The central storage facility runs at optimal efficiency using fewer but more powerful machines," he says.

This effective creation of an economy of scale can then be easily monitored for its energy use and cost efficiency, while the end users get automatic updates and maintenance.

Francois Zimmermann, UK chief technology officer at Hitachi Data Systems, says: "Banks need to remove the silos of data that exist and ensure that users have the ability to access data on the server from anywhere or any device. One of the best ways to achieve this is by virtualising servers to streamline information while providing businesses with greater agility and sustainable IT."

But without effective monitoring virtualisation can create a false economy. Indeed if the impression is that by using a virtual server less space and thus less energy is being consumed the temptation is to see that as a fait accompli, rather than to acknowledge the need to actively manage what is being stored virtually and whether it is still useful or relevant.

"Initially, virtualisation does allow for server consolidation, 300 virtual servers could be stored on just 20 physical servers. But without effective monitoring there is the risk that the virtualised services become redundant and are left to fester - thus taking up space and energy," says Mark Blackburn, chief technologist at IT optimisation specialist 1E. "This is all the more so when you consider that there is not a large capital outlay for setting up a virtual server - and that maintenance costs are less than $1,000 a year compared to $4,000 per annum. With relatively swift churn of IT personnel too, it becomes hard to keep on top of this sort of thing. Some 15% of servers are not actually doing anything useful and could be switched off."

Monitoring the servers also means that their use over a given timeframe can be maximised. For example a virtualised machine that is monitored uses only what is needs at any one time and can redirect its efforts when required. For example, machines that sit idle at night can be used by servers to run updates or analytics. Making best use of servers in this way reduces energy use, frees up disc space, reduces cooling load and allows for a "drowsy" setting on a machine that is not doing much.

In this context using a third party software provider with either its own data centre capability or links to specialist data centres, is proving popular.

Indeed where control and security is key, having the cloud-like structure that a virtualised service provides is attractive, but without the risk of having private data in a public cloud or indeed going to the hassle and expense of setting up a private cloud. Huge third party data centres solve the privacy, the economy of scale and the cost and sustainability issues in one fell swoop.

Graeme Creasy, director of operations at Interxion, a specialist col-location datacentre firm, says: "We can bring an economy of scale to the table - and for financial institutions one of our key selling points is having so many data centres that we can offer proximity to the banks that are trading on a millisecond."

The formation of Lloyds Banking Group in 2009 is a good example. It transformed its data centre using a third-party virtual storage platform. The legacy modular storage model could not support the new enterprise environment that had effectively doubled in size and needed about 50 million transactions uploading every night, the capacity to allow for data mining from most of the bank's middle office systems as well as replication, in its entirety, to a disaster recovery site. Lloyds needed a storage platform that would guarantee availability and optimal performance as well as the ability to scale out when required. At the same time the bank needed to ensure it could provision resources quickly to the parts of the businesses when required.

But can this sort of economy of scale also translate into sustainability on a worthwhile scale? Fortunately where data centres are concerned there are lots of small things to be done that all add up to a significant impact of energy consumption. The biggest thing is increasing the temperature within the data centre which reduces the need to cool. IBM recently recommended turning off CRAC - computer room air conditioning - units where no IT equipment load requires cooling. Where cooling is required, the minimum acceptable temperature can be raised a few degrees and the temperature can then be monitored using variable frequency drives (VFD). Cooling applied only when required, as opposed to fans working at 100% all the time.

Creasy says: "The savings to be made with VFD are immense, something in the region of 30% per annum in a room of 100,000 square foot. Another thing to watch out for is when introducing new cabling holes are cut in the floor but they are not routinely covered up and this then leaks pressurised air from under the floor into the room - wasting it.  The holes then need covering to minimise cool air loss."

Another seemingly common-sense way in which energy use can be reduced is by cold aisle containment. This is akin to placing large and small servers together- in a similar way to juggling different things in an oven according to whether they need to be hotter or cooler.
Ultimately, be it through data management, systems monitoring or data centre management, the key is that there are a lot of small steps that can be taken that really add up when placed together. And the happy combination of regulation, consumer interest, and the need to satisfy shareholders, means that operating sustainably is high up on the agenda of its own accord, supported by the broader climate of banks needing to make best use of systems and structures already in place.

Case Studies: Energy use/switching things off

It's not just at strategic system level that changes can be made. Encouraging change at the individual level also makes a difference. VocaLink, for example, reduced its printing by 45% in 2010 simply by making individuals sign into printers and making the default print out black and white and double sided.

At company level energy supply from renewable sources is increasingly coming into play and is often demanded at tender stage from third-parties such as large data warehouses as firms seek to green their entire supply chain.

But one of the easiest ways to reduce energy consumption is by making sure that individual machines are powered down when not in use - thus also freeing up bandwidth to be used elsewhere within a network.

1E's clients have seen particular success with this. At HSBC, a project over 2009 and 2010 meant the total number of global PC shutdowns on weekdays and weekends in 2009 saved $1,090,180 in energy costs, 201,885,139 hours of computing power, 7,267,865 kWh of energy and 3,125,182 kg of CO2 emissions.

And a 2011 project at New Zealand's Kiwibank enjoyed a 163,800 kWh reduction in energy use, NZ$24,570 projected energy cost savings and 32 tonnes of reduced CO2 emissions by monitoring and reducing computer energy usage. The bank had found over 50% of computers were being left on overnight and worked to stop this as well as patching computers out of business hours and thus reducing peak load during business hours.