The same fiber-optic cable that can read temperature all along a wellbore can also be made to listen. Distributed acoustic sensing uses the fiber as an array of thousands of virtual microphones, detecting the tiny strains that sound and vibration impose on the glass at every point along its length. That means an operator can hear, depth by depth, the noise of fluid entering the well, sand scouring the pipe, gas hissing through a leak, or a hydraulic fracture growing during a stimulation. The catch is that acoustic sensing samples so fast and at so many points that its raw data rate dwarfs even a temperature fiber, which is why serious DAS deployments push much of the processing to the edge before anything reaches a control system.
Distributed Acoustic Sensing (DAS) in one line: Distributed acoustic sensing, or DAS, uses an optical fiber run into the well as a dense array of acoustic and vibration sensors, detecting sound at thousands of points along its length rather than at one location. It works by sending laser pulses down the fiber and measuring how backscattered light shifts as passing acoustic energy stretches the glass, which yields a signal at every depth. In wells it is used for flow allocation, sand and leak detection, and monitoring hydraulic fracturing, and its very high data rate makes edge processing essential.
DAS uses the same kind of optical fiber as a temperature system, often the very same cable, but it interrogates it in a way tuned to motion rather than heat. The surface instrument fires rapid laser pulses down the fiber and watches the faint light that scatters back from natural imperfections spread continuously through the glass, a process called Rayleigh backscattering. When a sound wave or vibration reaches a section of fiber, it stretches and compresses the glass by a minuscule amount, and that strain slightly shifts the phase of the light scattering back from that section. By measuring those phase changes, the instrument recovers the acoustic signal at every point along the fiber.
As with a temperature fiber, the return time of the light tells the instrument which depth each measurement came from, so the fiber acts as a long line of independent sensors, one for every short segment of its length. The crucial difference is the timescale. Sound is a fast phenomenon, so DAS interrogates the fiber thousands of times per second at every depth, capturing waveforms rather than slow trends. The output is not a single temperature curve but effectively a dense grid of acoustic channels, each recording a rapidly varying signal, which is what makes the fiber behave like an array of microphones distributed the entire length of the well.
This acoustic view complements the thermal view rather than replacing it. Temperature sensing is slow and steady and excels at showing where fluid is entering or leaving over minutes and hours; acoustic sensing is fast and responsive and excels at catching events, the burst of noise when a zone starts flowing, the roar of a frac, the hiss of a leak. Because a single fiber can carry both interrogators, many wells run DTS and DAS together to get the complete picture, the temperature profile telling the operator where slowly, and the acoustic profile telling them what is happening right now and how loudly. The two measurements share a cable but answer different questions.
One of the most valuable uses of DAS is figuring out where flow is coming from along a well. Fluid moving through perforations and up the wellbore generates broadband noise, and turbulent inflow at a producing zone has an acoustic signature distinct from quiet pipe. By analyzing the noise depth by depth, engineers can identify which perforation clusters or which zones are actually contributing and estimate their relative flow, giving a flow allocation across a long horizontal completion without running a wireline tool. Because the fiber is permanent, this allocation can be repeated as the well ages and its zones change, something a one-time production log cannot match.
The acoustic signature is also how DAS catches sand and leaks. Sand entrained in the flow scours the inside of the tubing or strikes it at bends and restrictions, producing a characteristic high-frequency noise that rises sharply when sand production starts, so a DAS system can flag the onset of sanding before it erodes equipment. Leaks announce themselves the same way: fluid or gas forcing through a casing breach, a bad connection, or a channel behind pipe generates noise at the point of escape, and because the fiber covers the whole well, the leak's depth is part of the signal. That location information is what makes DAS a genuine integrity tool rather than just a flow diagnostic.
In hydraulic fracturing, DAS has become a core monitoring method. During a stimulation, the fiber can watch fluid leave each perforation cluster and see how evenly a stage is being treated, revealing whether some clusters are taking most of the fluid while others are starved. Fibered wells are also used to listen to fracturing in a neighboring well, detecting the arrival of a frac hit as the treatment propagates across the space between wells. This near-real-time view of where the energy is going lets a completions engineer judge stage efficiency and cluster uniformity while the job is running, feeding decisions about diverter use and stage design in a way that after-the-fact analysis cannot.
DAS produces more data than almost anything else on a well site, and by a wide margin. A fiber sampled thousands of times per second at thousands of depths generates a torrent of acoustic samples, easily amounting to enormous volumes of raw data every hour. There is no realistic path by which that raw waveform stream is sent over a field network to a central historian or a cloud SCADA system in its native form; the bandwidth off a remote pad and the storage at the other end would both be overwhelmed almost immediately. Unlike a slow temperature profile that can be trimmed to a few tags, DAS is fundamentally a high-rate signal-processing problem before it is a monitoring one.
For that reason, serious DAS deployments do the heavy lifting at the edge, on hardware sitting at or very near the wellsite. On-site processors take the raw acoustic array and extract the features that actually matter, converting waveforms into results such as a flow allocation per zone, a sand-detection alarm state, a leak location and severity, or a frac-monitoring summary for each stage. The full raw data may be recorded locally for later detailed analysis or discarded after processing, while only the compact, meaningful outputs are forwarded onward. This edge-first architecture is not optional for DAS; it is the only way to reconcile the sensor's data rate with the realities of field networks.
Once the edge has reduced the fiber's output to those meaningful signals, a cloud monitoring platform such as Merobix can carry them like any other well measurement. The distilled outputs, a sanding alarm, a leak flag with a depth, an updated flow allocation, sit alongside the well's pressures, temperatures, and rates so an operator sees the acoustic insight in full context and gets notified when something changes. The division of labor is the same one that makes any high-rate sensor practical in the field: the fiber and its interrogator sense at the raw rate, edge processing turns that into decisions and alarms, and the cloud layer delivers those results to the people who run the field without ever needing to move the raw waveforms.
DTS measures temperature slowly along the fiber, while DAS measures fast acoustic and vibration signals along the same fiber. DTS is best for showing where fluid enters or leaves over minutes and hours, and DAS is best for catching events like inflow noise, sand, leaks, and fracturing in near real time. Many wells run both interrogators on a single cable because the temperature and acoustic views answer complementary questions.
Fluid or gas forcing through a casing breach, a failed connection, or a channel behind pipe generates noise at the point where it escapes. DAS senses that noise at whatever depth it occurs because the fiber covers the entire wellbore, so the acoustic signature reveals both that a leak exists and where along the well it is. That location information is what makes DAS a practical integrity tool rather than just a flow diagnostic.
DAS samples the fiber thousands of times per second at thousands of depths, producing an enormous raw data rate that would overwhelm both the network off a remote site and any central historian. The standard solution is to process the raw acoustic data at the edge, near the wellsite, and extract compact results such as flow allocation, sand alarms, and leak locations. Only those distilled outputs are sent to the control system or cloud platform, not the raw waveforms.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.