Pretty much all the risk will come from things the team has not done before (and every project has some aspect of that, unless you're experts in the particular field, in which case it isn't a 'project' so much as a 'job' - rolling out SAP in yet another enterprise, for example).
So, assuming you've got some things that the team has never done, first you need to find those things. Here's my step-by-step.
Find the stakeholders and their goals. There'll be some championing stakeholder for the project, but also a bunch of people who need to be kept happy in order to make it succeed - legal, audit, etc.
Divide the project into capabilities. The capabilities will be those that can help the business, or users, to achieve the stakeholder's goals. For instance, a software project might provide the capability to book a certain type of trade, calculate a particular aspect of risk, limit exposure to any one customer. A mobile phone has the capability to receive calls. Note that I am not referring to any concrete features here - the capability could be delivered through filling in paper forms, for instance.
Ask the team to identify which things they've never done before. For instance, while the ability to make and receive calls on a mobile is pretty standard, maybe the particular type of camera you're using is new. You can also look at which stakeholders have never been considered before. It could be that you're going to be delivering a capability in a new way - with a new UI, etc. - in which case, you probably have a new capability like "do X faster" as part of that.
Identify which things have been done before by someone else. These are things which require expertise you simply don't have. They're lower risk than the other things which have never been done before by anyone.
Identify things which have never been done by anyone. This is where your highest risks are.
By identifying things that the team has never done, you're bounding your ignorance. Brainstorming and other techniques tend to bring up things which are known or have been considered, but the biggest risks are in the things which haven't been considered. Pretty much those happen in the places where new stuff is occurring.
For good reading around this, I recommend Waltzing with Bears, and Dan North's post on Deliberate Discovery, which is written from a software perspective but IMO has some lessons for all walks of life.