Energy storage (ES) is vital in electricity markets and its operation is currently forced on single market service. However

ES could get more profits by participating in multiple market services. Thus

the potential of ES is undervalued

and ES cannot get sufficient incentives. In addition

the benefits and risks vary significantly over different markets

which causes severe challenges for ES to manage its capacity optimally over its daily cycle. Portfolio Theory

an optimal trading tool to maximise expected return while minimising corresponding risk

is introduced to quantify the ES capacity allocation over multiple markets. This paper models expected returns and risks of ES over the energy arbitrage market

enhanced frequency response market

and distribution network operator's (DNO's) market. Corresponding to risk over markets

risk aversion level of the storage is evaluated based on the Cumulative Prospect Theory

which is depicted as its indifference curve. The optimum portfolio is calculated by the intersection between the efficient frontier and indifference curve. Then

the ES operation strategy is designed via Robust Optimisation to ensure maximum profit under market price signals uncertainty. Results show the proposed method models risk aversion level of ES more accurately. It also provides an efficient capacity sharing model and daily operation model for ES participating in multiple market services to gain maximum economic return at minimum risk under uncertainty.