A substantial and problematic ribbon lengthening is sometime observed in PV modules exposed to thermal cycling. In large amplitude thermal cycling, it is systematically observed with standard architectures and materials. To our knowledge, there is no available explanation of this effect in the literature. In the present work, a ribbon lengthening mechanism is derived from a simple three layers model. This mechanism relies on the relative thermal expansion of the copper ribbons with respect to the silicon cell, the elasto-plastic behavior of the ribbons and the temperature dependent visco-plastic behavior of the solder. It neglects the existence of the front/back sheet and the encapsulant. An analytical derivation provides a lengthening criterion which depends on the cycling thermal amplitude. Using the Finite Element Method (FEM), a lengthening rate is predicted. Then, it is measured in single cell strings with different thermal amplitudes by indirect means. Finally, the lengthening rate is also measured in single cell laminated modules using local image correlation. The three layers model predictions are in good agreement with the indirect measurements. Nonetheless, direct measurements results are not consistent with it. From these observations and others, it is inferred that the encapsulant comes into play after the solder failure. Its role is further discussed in the Part II of this work.