Systematic effects in dual-beam, differential, radio observations of extended objects are discussed in the context of the One Centimeter Receiver Array (OCRA). We use simulated samples of Sunyaev-Zel'dovich (SZ) galaxy clusters at low (z < 0.4) and intermediate (0.4 < z < 1.0) redshifts to study the implications of operating at a single frequency (30 GHz) on the accuracy of extracting SZ flux densities and of reconstructing comptonization parameters with OCRA. We analyze dependences on cluster mass, redshift, observation strategy, and telescope pointing accuracy. Using Planck data to make primary cosmic microwave background (CMB) templates, we test the feasibility of mitigating CMB confusion effects in observations of SZ profiles at angular scales larger than the separation of the receiver beams.