The 34-minute power outage at Superbowl XLVII in January 2013 was traced to the operation of a relay. With 108 million Americans looking on and the event being broadcast to over 180 countries worldwide, it was a timely reminder of the potential for major disruption to occur due to the failure of electrical power equipment.
Such outages, as they are known in the power industry, can come about as a result of accidental or deliberate attempts to disrupt the systems involved in managing the networks. On a wider scale, the huge blackout that affected the North East of America in 2003 was quickly attributed to a cyber-attack.
Rumours quickly circulated suggesting that the attack had been caused by the MSBLASTER Worm that was causing problems in some systems at the time. It would seem that that event was not the result of a state trying to disable power systems in the US.
However, the after effects of the event are not lost on current military leaders, who have ventured to suggest that if a state could have been proven to have tried to take down power networks that would provide the Casus Belli for war.
Despite all the hyperbole associated with the idea of cyber warfare, in practice little of what has happened so far in cyber space can be described as warfare. There are clearly many examples of cyber espionage that have appeared in the media.
National governments are constantly warning major international companies about the potential loss of their intellectual property. Some studies have gone so far as to suggest that the total losses already to such acts of industrial espionage in cyber space runs into billions of pounds.
Whilst that is of concern to ministers, the one nightmare that has yet to be fully realised is the aftermath of an attack on power distribution networks. Small scale events have seen specific parts of the power system apparently targeted.
In late 2011, in Springfield in the US, a pump at a public water utility was made to cycle its operations so frequently that it broke down. Suggestions quickly circulated in the media that the SCADA system controlling the pump had been remotely accessed, with an IP address in Russia being touted as a potential source of the attack. This is just one example of what many believe to be an insidious pattern of low level attacks.
For some these herald the way in which cyber warfare may manifest itself. A foreign adversary may be unable to physically attack a country due to a lack of military equipment, but using cyber-space it can still cause harm and disrupt vital utilities.
During the Kosovo crisis, reports emerged of Serbian hackers attacking a number of NATO countries critical national infrastructure. They returned to their attack as French and British jets operated over the skies in Libya in 2011.
Where a potential adversary lacks the ability to physically attack a country they can chose to manoeuvre in cyber-space. Recent attacks against Saudi oil production facilities have been suggested as one example of a possible state-led intervention.
Attacks against the New York Times have also been suggested as coming from China. Iran smarting from the cyber-attack that disrupted its nuclear research programme using the STUXNET Worm is clearly actively seeking to add cyber capabilities to its military arsenal.
Utility companies are specifically vulnerable in this dynamic environment. They are constantly under media and public pressure whenever the cost of electricity or gas increases. Keeping costs down forces them to look at ways of increasing the efficiency of their operations.
Supervisory Control and Data Acquisition (SCADA) systems are an essential element of the day-to-day operations of major power distribution companies. But as the Iranians found out to their cost, when the STUXNET virus was introduced into their systems they can be very vulnerable. This is especially the case if wireless systems are deployed.
Historically, SCADA systems have operated on dedicated networks that are relatively secure. Using fixed land-lines made it physically difficult for hackers to gain remote access to systems. This has meant that the building blocks of SCADA systems have not been particularly secure. Password protection was at best basic, but sufficient for the environment in which it operated.
Such fixed networks also provided guarantees with respect to the latency of message passing. As communications providers moved to delivering service-based networks, problems arose with message latency. This has forced organisations such as the UK’s National Grid to look around for alternative ways of communicating data between its various SCADA systems.
In looking for solutions, it was natural for the National Grid to look carefully at the potential of wireless systems. Operated as point-to-point networks their latency characteristics were predictable. Once major users of networks switched over to service-based approaches their finely tuned SCADA systems started to experience problems with message delays. In the past, SCADA systems have also been able to rely upon specific propriety communications protocols developed by specific manufacturers. This also ensured that where radio systems were employed, anyone seeking to gain access to the system had to be both within radio range and able to generate the propriety protocols.
This gave SCADA operators a level of default security that many did not fully appreciate. But over time pressure from SCADA systems customers has forced many suppliers to develop communications systems with a standard protocol. This has made it easier for such systems to be attacked.
As 3G and 4G networks proliferate, the utility companies are turning to wireless networks to create communication’s backbones for their SCADA systems. Moving to a wireless-based system offers many advantages.
Not least amongst these are the benefits gained from the deployment of advanced technologies that require less power. Equally important is the benefit of using the networks to upload software revisions to the SCADA data logging nodes.
In the past this has been a time-consuming and often cumbersome process. With the additional bandwidth available from wireless systems such operations are performed quickly. Several technical solutions are also available to enable the wireless systems to be deployed.
Satellite-based networks, private radio networks or cellular systems all offer slightly different benefits depending upon the geography of the systems being monitored. But the move to wireless based solutions has to be weighed against the basic security provisions in SCADA networks. If a determined hacker can get into the line of sight of the wireless network, they might be able to gain access to the network.
Many existing remote data logging systems are served by fixed landlines, whose cost of maintenance and installation is increasing. For utility companies, wireless networks can often be co-located with their own facilities. This is the case with the National Grid in the UK.
The fear is that as the utility companies move to new SCADA systems, they will not pay sufficient attention to the need to ensure their systems are secure. For terrorists, the potential to reach out across the internet from a remote location via mobile systems to disrupt a major utility in Europe or America is an attractive one.
With an estimated 90% of the world’s population now within range of a mobile service, the danger has increased dramatically in the last few years.
Whilst the power outage at the Superbowl was temporary and easily remedied, a state-based or terrorist attack against vulnerable SCADA systems could produce a very different and much more spectacular outcome.