With the promotion of “carbon neutrality” and “carbon allowances trading” policies

the importance of evaluating the carbon emission reduction efficiency of power plants is becoming increasingly prominent. Carbon emission reduction efficiency of power plants is affected by capital

labor

energy consumption and other factors

among which the level of carbon management is also an important input factor. However

because there are so many things related to carbon management

the cost is difficult to quantify because of great uncertainty. Therefore

this study introduces the interval type-2 fuzzy number(IT2 FNS) to deal with the linguistic variables of experts' assessment of the carbon management level of power generation enterprises

and constructs the super-efficiency slack based model(SBM) with interval type-2 fuzzy number and undesired outputs. The model is divided into four sub-models for solving and sorting by using α-cut and optimistic and pessimistic estimation methods. The data of 17 power plants in four provinces of G Energy Group is selected to evaluate the overall carbon emission reduction efficiency level of the selected plants. The results show that

different power generation enterprises have different sensitivities to the uncertainty of carbon management input. Power generation enterprises in Jiangsu province generally have the highest carbon emission reduction efficiency

then is that in Guangdong

Liaoning and Shaanxi

and the carbon emission reduction efficiency levels of the four provinces are uneven.