Normally no electromagnetic interference between the power cables and the communication cables, because each has a different frequency. For the power cables the frequencies used are 50 Hz in Europe or 60 Hz in America (low frequencies), and for power supply cables are high frequencies (more than 900 MHz).
There may (and has been and still is) be a lot of interference (coupling) between power cables an signal cables.Thats one of the reasons why people like to use optical fibers for communications where some of the first practical tests were done in subway installations.The argument that the frequency range is different is not really true since there may be all sort of transient on the power cables which have spectral components going into the communication range or make problems via intermodulation. With proper filtering however , communications can (and are) transmitted via power lines
Although my domain is EMI & EMC issues at rather high frequencies, in my humble opinion the main source of interference in this case will be magnetic field associated with the power cables. Thus distance will be a very important factor. Moreover, the interference could actually be affecting the terminal circuitry more than the cables themselves. If you tried shielding these cables and still get interference I suggest try using magnetic field shields like mu-metal etc.
The inductive interference caused by ac electric traction lines to nearby telecommunication cables has been investigated in this paper: K. J. Satsios, D. P. Labridis and P. S. Dokopoulos, "Inductive interference caused to telecommunication cables by nearby AC electric traction lines. Measurements and FEM calculations," in IEEE Transactions on Power Delivery, vol. 14, no. 2, pp. 588-594, April 1999. In this article there is useful conclusions concerning the influence of the separation distance between the electric traction line and the
telecommunication cable, of the earth resistivity and of the mitigation wires on the inductive interference levels have been obtained, by using a FEM parametric study.
Other important references:
International Telecommilnication Union : "Directives Concerning the Protection of Telecommunication Lines against Harmful Effects from Electric Power and Electrified Railway Lines", vol. I and vol. 11, Geneva 1989.
J. R. Abbott, "HVDC System Simplified Method of Harmonic Coupling Calculation and Testing", /IEEE Transactions on Power Delivery, PWRD-3, April 1988, pp. 839-844.
Short answer is that Cross coupling is a well known phenomenon that always exists between neighbouring cables. Now, this coupling may be of two types, Capacitive and Magnetic. Capacitive coupling prevails when the Impedance of load on power cable is high or there are high voltages on power line. While magnetic coupling is dominant when load Impedance is low or there is hight current on power line.
Now, the question is that how these couplings exist between power cables and data cables. The switching loads produce strong transients on the power cables. These transients could have wide band frequency content, up to several mega hertz. Thus, these coupled transient bursts can distort several packets of digital data.
Magnetic coupling can be reduced by physically increasing the gap between cables. While, capacitive coupling can be reduced by using a good conductive shield
The power cables are mainly sources of magnetic fields which can induce voltage on the data cables. However, the m-field decays quickly as the the parallel separation between data and power cables increases. It can also be filtered easily since it is 50 Hz noisy signal. On the other hand, it is not easy to shield the magnetic field. Thus, the layout of power cable using cancellation effect can reduce tremendously the induce voltage such as balancing the current flow in 3-phase cable. Please refer to EN 50174 standard for more details and the other EMC books as well.