Inside every frac stage the perforations are not spread evenly along the casing - they are grouped into a few short bursts, and each of those groups is a perforation cluster. Clusters are where the treatment actually enters the rock, so how many there are, how they are spaced, and how evenly they each take fluid decide how well a stage works. This guide explains what a cluster is, how it differs from a stage, how limited-entry design tries to balance flow among clusters, and how engineers infer what each cluster took.
Perforation Cluster in one line: A perforation cluster is a short group of perforations - typically a few holes over a span of about a foot or two - shot into the casing within a frac stage. A single stage usually contains several clusters, and the stage's fracturing treatment is divided among them, so each cluster is intended to initiate its own fracture. How many clusters a stage has, how far apart they sit, and how equally they each accept fluid are central questions of completion design.
A frac stage is the entire isolated interval of the lateral that is pumped as one event, and a perforation cluster is one of the small groups of perforations inside that interval. If you picture a stage as a stretch of casing sealed off at the bottom by a plug, the clusters are the handful of short perforated windows spaced along that stretch. When the fracturing spread pumps the stage, the fluid and proppant leave the casing through those cluster windows, so the number of clusters sets how many entry points the stage has into the formation.
The reason a stage is subdivided into several clusters rather than perforated as one long window is that a single continuous set of perforations tends to accept fluid only where the rock is weakest, concentrating the entire treatment into one spot. Splitting the stage into distinct clusters spaced apart is meant to seed several fractures across the interval instead of one. Cluster spacing - the distance between clusters - controls how finely the stage is subdivided, and tighter cluster spacing paired with shorter stages is the general direction completion designs have moved to contact more of the rock per foot of lateral.
So the hierarchy runs lateral, then stage, then cluster. An engineer chooses a lateral length, divides it into stages, and divides each stage into clusters. The stage is the unit of isolation and of pumping, while the cluster is the unit at which fractures actually initiate. Getting more of the reservoir producing means not just placing enough stages and clusters but getting each cluster to take its share of the treatment, which is a harder problem than it first appears.
Even within one stage, clusters do not naturally take equal amounts of fluid. The cluster in the weakest or lowest-stress rock will happily accept far more than its neighbors, so a stage can end up dumping most of its proppant into one or two clusters while the rest are starved. Limited-entry perforating is the technique used to fight this. By shooting a deliberately restricted number of perforation holes of a chosen diameter, engineers create a pressure drop across the perforations themselves that is large enough to force fluid to spread more evenly among the clusters rather than all rushing into the easiest one.
Extreme limited entry pushes this idea further, using very few, carefully sized holes per cluster so the perforation friction dominates and dictates how fluid distributes. The design also considers perforation phasing - the angular arrangement of holes around the casing - and the erosion the holes undergo as abrasive proppant passes through them, since erosion enlarges the holes and gradually erodes the very pressure drop the design relies on. Getting these details right is what determines whether the treatment is shared across the clusters or wasted on a few.
Cluster efficiency is the metric that captures the outcome: the fraction of clusters that actually took fluid and contributed a fracture, as opposed to sitting idle. A stage where every cluster breaks down and accepts proppant is efficient; a stage where half the clusters never open is not, and that half of the perforated rock produces little or nothing. Because starved clusters represent reservoir that was paid to be perforated but never stimulated, cluster efficiency is one of the most watched outcomes of a completion design, and improving it is a major reason limited-entry and diversion techniques exist.
The practical difficulty is that a stage is pumped down a single string, so the surface measurement is one combined treating pressure and one combined rate for the whole stage - there is no direct flowmeter on each cluster thousands of feet downhole. Engineers therefore infer how fluid split among clusters from the signals they do have. The treating-pressure response as the stage breaks down and takes proppant carries clues, and post-job techniques that examine how the perforations eroded can indicate which clusters saw the most slurry. Increasingly, fiber-optic sensing and distributed acoustic sensing, or DAS, are run to listen to flow along the wellbore and estimate the per-cluster distribution more directly.
All of these diagnostics depend on high-quality, high-resolution data captured while the stage is pumped and preserved afterward for analysis. This is where field data acquisition and cloud monitoring matter. A cloud SCADA platform such as Merobix can ingest the treating pressure, rate, and proppant channels streaming off location and tag them by stage, giving engineers the continuous, well-organized time series they need to interpret cluster behavior. When fiber or DAS data is also available, having the treating record aligned with it in one accessible system makes the per-cluster picture far easier to assemble.
The value of getting this right is direct: if engineers can tell which clusters were starved on which stages, they can adjust perforation counts, cluster spacing, limited-entry design, or diversion on the next well to even out the distribution. That feedback loop only works if the data from every stage is captured cleanly and kept where completion engineers can reach it. Streaming the treating channels into a monitoring platform, rather than leaving them stranded in a data van, is what turns a one-time job record into an input for continuously improving cluster efficiency across a program.
It depends on the completion design, but a stage commonly contains several clusters spaced along the interval rather than one long set of perforations. The count has generally risen as operators tighten cluster spacing to contact more rock per foot of lateral. The exact number is chosen for each well based on stage length, rock properties, and the limited-entry design being used.
Limited-entry perforating uses a deliberately restricted number and size of perforation holes so that friction across the perforations creates a large pressure drop. That pressure drop forces fracturing fluid to spread more evenly among the clusters in a stage instead of all rushing into the easiest one. Extreme limited entry pushes this further with very few, carefully sized holes per cluster.
Cluster efficiency is the fraction of perforation clusters in a stage that actually took fluid and initiated a fracture, rather than sitting idle. A high cluster efficiency means the treatment was shared across the perforated interval, while a low one means much of the rock that was perforated never got stimulated. It is a key measure of how well a completion design distributes the treatment.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.