Measure LPG (LP Gas) level using an ultrasonic sensor. Which sensor

I didn't look at the sensors, but in general RF can be a problem when it gets into low level analog circuitry not intended for those frequencies. The RF you send should be of little consequence because this should be intermittent right after you take a reading, with a bunch of time before the next reading.

What you should really be worrying about is ambient RF like local radio stations and noise from nearby equipment. Mechanically commutated electric motors can be substatial RF noise sources, for example.

Fortunately, this is relatively easy to deal with. Since your measured signals are much much lower in frequency, the RF noise can be easily attenuated by simple passive low pass filters. At RF frequencies, you can't rely on common mode rejection ratio and other parameters of analog circuitry. Frequencies above what the opamps and the like are intended to handle can end up getting rectified and the resulting baseband AM modulation value ending up as noise on your intended signal. That is much harder to separate out since it can be in the same frequency range as your signal.

The solution is to not let frequencies above your intended signal get into the active circuitry in the first place. One or two poles of R-C filtering is cheap and simple. If you can't afford to add much impedance, then maybe one or two ferrite chip inductors followed by caps to ground will do. Fortunately, the ratio of these RF noise frequencies to the highest end of the pass band is large, so there is a lot of room for simple passive filtering.

The SRF05 sensor is designed to give a pulse indicative of the time it took to receive the first reflection of the US signal. Hence when a nearby object is present a more distant object will have no effect at all on the 'official' output.

Somewhere on the PCB you can tap into the raw received signal. You could process this yourself. If the more distant object produces a US reflection it will be observable in this signal.