Understanding the fine print on contracts for data center space
Key issues: awareness will help speed the deal
If you are involved in data center projects, an understanding of the key issues which arise from the underlying contractual documentation can assist you to execute deals faster, keep costs lower, reduce liability and future-proof your business. Data center deals vary in scale and complexity and have different legal structures. The issues raised below envisage a customer contracting with a data center operator for the delivery and use of data center space.
1. Basis of the customer’s occupation. This will vary between jurisdictions and it is important to understand the implications. In the UK, for example, occupation will be on the basis of a lease or a licence and there are fundamental differences between the two, with a lease giving the customer greater control, security of occupation and protection in the event of operator insolvency (important factors when infrastructure is supporting mission -critical data) but involving additional cost, including property tax. Flexibility will also be important – including the customer’s ability to install/remove equipment, make other alterations or assign or terminate the contract – and the parties may have competing interests which need to be balanced.
2. Design and delivery. Control over design and specification changes can cause great debate and the parties’ rights and responsibilities need to be clearly stated. Timing and measurement will also be critical and the customer may look for specific damages and ultimately the right to terminate if targets are missed. The ‘when/how/who’ of commissioning and sign-off of works needs to be addressed, with a well-advised customer wanting input into these processes.
3. Construction documentation. The underlying building contract and the professional appointments are key when significant works are involved, with greater importance being attached to the mechanical and electrical installations than on other construction projects. The operator will want to ensure its design/construction obligations to customers are passed down to its construction team wherever possible. A customer should consider requiring the ability to approve the identity of that team and obtaining duties of care from them.
4. Site specifics. Site due diligence is important to establish fundamentals such as rights of access, availability of power and requisite planning consents. Connectivity is also key and customers should ensure they have sufficient rights (and obligations on the operator) to bring in new telecommunications providers.
5. Environmental liability and taxes. The parties need to consider and address risk of potential environmental liability in respect of the site. Consider also relevant carbon emissions trading schemes under which the operator could be primarily responsible for the customer’s emissions, and how associated costs should be re-charged in this situation.
6. Clarity of scope. The contract needs to set out clearly what the operator is obliged to provide in terms of services and how related obligations (such as replacing fuel in generators) work. It is also important that the documentation allows the level of flexibility required throughout the life of the contract.
7. Clarity of price. An obvious point, but the price the customer is paying needs to be clear. Where payments are based on actual power consumed then this is relatively simple, but where they are based on assumed power consumption – perhaps by reference to number of racks installed – then some detailed thinking is needed. Is a rack treated as live once installed or only once drawing power, for example? Is there to be flexibility for the customer to draw or reserve additional power, and how does this affect the operator’s cost and revenue projections?
8. Clarity of service levels. Operators typically commit to service levels around main KPIs such as power, humidity and temperature, which are tied to a service credit mechanism. Various ‘what if’ scenarios need to be considered, a good example being where one service level failure – for example power outage – naturally leads to other service level failures: should multiple service credits arise or should they be limited to the primary failure? And for how long does a service level need to be breached for a credit to arise? A short lived temperature spike may have no major impact, but from the customer’s perspective repeated failures caused by the same problem should not count only as a single failure as otherwise the operator has less direct incentive to fix the issue quickly.
9. Effectiveness of remedies. What if things go wrong? The customer needs the ability to control risk going forward, particularly where the operator has committed a number of minor failures that undermine confidence but don’t necessarily give the customer any significant leverage under the agreement. In these situations a customer may have the right to exercise a degree of management oversight. A more powerful remedy is for the customer to step in and actually perform the services, although this is not always achievable in practice, particularly where plant supports multiple customers.
10. Liability. The parties need to strike a bargain on liability that allocates risk appropriately. Typically the parties will agree exclusions of certain types of loss (e.g. loss of profits) and apply caps on liability that reflect the insurance cover available to the parties.
The extent to which any of these issues are relevant will depend on the circumstances and the nature of the deal, but whatever the circumstances the parties each need to consider the key risks (the ‘what if’ scenarios) and ensure they have measures in place to mitigate them.