Problem no: BPO11
Related topics:
- TEP 4235 - Energy Management in Buildings, see
Background
Energy service companies (ESCOs) have a program to help their clients optimize the energy application of office blocks. With the optimization, clients can save large amounts of expenses. According to the program contract, ESCOs need to reduce half of the energy fee. Then they can get 50% of the saved expense as rewards annually for ten years. Theoretically, the optimization design can achieve the goal agreed in the contract. However, the optimization effects can be influenced by various factors from technical, environmental, climatic and managerial perspectives. For example, if the temperature becomes much colder than previous years, energy fee may only reduce 40% instead of the expected 50%, even if an excellent energy optimization has been completed and the saving goal should have been achieved. Besides such uncontrollable climatic factor, the problem can also be caused by poor optimization technique, which is considered as the fault of ESCOs. In contrast, clients’ behavior, such as adding high-power facilities in the block, can also lead to undesired saving effect. In this case, clients should be responsible for the unachieved goal. Disputes between ESCOs and their clients often occur when the agreed optimization effect is not fulfilled. In such challenging situation, a tool is needed to analyze the causes of goal failure and estimate the responsibility of ESCOs and clients in a scientific way. This tool can make the two parties understand what goes wrong and why it happens, which benefits the solution of the dispute. Moreover, the tool can have another function, helping ESCOs predict the risk of goal failure before signing contracts considering potential influential factors. For the blocks with different risk levels, ESCOs can provide different offers to avoid disputes.
Brief description of the assignment:
(1) Assessment model development. The influential factors of energy consumption are identified using fault tree model (FT). Since some influential factors have interaction, which may change their effects on energy consumption, the dependency relationship between these factors are represented using Bayesian network (BN). These two models can be used to predict the risk in advance and analyze the causes given goal failure. The model structures can be decided first, and then relationships between variables need to estimate. This could be achieved in the second step.
(2) Effect simulation of an influential factor on its dependent counterpart and consumed energy. The quantitative relationships among different influential factors and consumed energy are needed to complete the development of BN. The quantitative relationships can be obtained based on simulation. For example, if temperature is identified as an influential factor, the increase amount of consumed energy caused by temperature decrease can be simulated using software. Then the quantitative relationships among temperature and consumed energy can established in BN.
Number of students on the assignment: 1
Contact person at IBM:
- Mohamed Hamdy (mohamed.hamdy@ntnu.no),
External partners:
- Guozheng Song (guozheng.song@ntnu.no), Department of Mechanical and Industrial Engineering.
- Hasan A. M. Hamdan, PhD research fellow, Department of Industrial Economics and Technology Management, Faculty of Economics and Management