The notion of effect in programming languages has evolved significantly since the works of Lucassen and Gifford – where an effect system tracks memory regions and enables the improvement of parallel execution – to the point where an algebraic characterisation of effects is proposed. In this work, our interest lies in an algebraic investigation of effects and how optimisable they are. Therefore, we propose a calculus, lambda genArt, that targets the development of generative art, which inherently demands effectful computations. We provide the semantics and type system of lambda genArt, alongside an effect algebra and a new parallel constructor. We implemented the calculus as a DSL for the Haskell programming language and introduced optimisations based on the effect information. This work is the first step towards the specification and implementation of a declarative functional language for generative art based on algebraic effects and handlers.