The limited availability of installation space downstream of valve-induced flow disturbances represents a critical metrological challenge, as ultrasonic flowmeters (UFM) exhibit high sensitivity to such disturbances. International standards currently overlook the impact of field installation on the accuracy of UFM. This paper presents and compares the influences of flow disturbances induced by a butterfly valve and a knife-gate valve on the performance of a clamp-on transit-time UFM. Measurements were performed at non-standard sections downstream of each valve by varying the angular position (α) of ultrasonic transducers in 30° increments around the pipe center. The tests were conducted for two distinct closings of the valve, corresponding to Reynolds numbers (Re) of 77,600 and 360,000. The main concern of this study is to evaluate and ensure that the measurements obtained in such non-standard sections are reliable and acceptable. This study further improves the accuracy of the UFM measurement based on a compensation criterion without maintaining the standard length of a pipeline. CFD models were created to provide insight the differences in installation effects. The CFD predictions show good agreement with the experimental results, with mean absolute percentage errors of 2.2% for the butterfly valve and 5.43% for the knife-gate valve. The results showed that, at the same Re, the transit-time signal stability and measurement accuracy differ noticeably between the two valve types. The findings showed that the spatial distribution of flow velocity strongly influences the performance of the UFM. The study revealed that at 9DN downstream of the disturbance element for higher Re, the correction factor (Cfac) does not exceed 1.02, where the measurement error (δ) and its repeatability uncertainty are within the permissible limits without the need to determine an optimal α. This study also provided certain values of α where the correction factor becomes minimal at distances below 9DN.