Modal filtration in the field of damage detection has many advantages, including: its autonomous operation (without the interaction of qualified staff), low computational cost and low sensitivity to changes in external conditions. However, the main drawback of this group of damage detection methods is its limited applicability to operational data. The modal filtration of the responses spectra (in place of FRFs), proposed in the literature, often does not give the expected results, working properly only for excitation in the form of white noise, or an ideal impulse. In other cases, for example in rotational machines, when in the response spectrum the rotational velocity harmonics dominate it can give wrong results. For such cases authors propose to use a new type of spatial filter, similar to modal filter with the difference that it has ability to filter the operational deflection shape components from the system response. Its application together with classical modal filter allows for damage detection using operational data with other type of excitation. The main assumption of the new spatial filter is the orthogonality of the filter coefficient vector to the operational deflection shape vector, it is then similar to the classical modal filter.