Towed streamer fundamentals

Part 4, Marine acquisition

Learning objectives

Draw the plan view of a marine 3D streamer survey: ship + sources + N streamers
Calculate total channel count from streamer length, group spacing, and N
Recognise why the near-offset gap (~300 m) is a fixed streamer limitation
Estimate shot interval along a sail line from vessel speed and listening time

A modern marine 3D survey is a single vessel towing between 4 and 14 streamers, kilometre-long oil-filled polymer cables containing hydrophone groups every 12.5 m. The ship tows a source array ~150 m behind the stern; the streamers fan out another 100-150 m behind the source. The first receiver group on every streamer is therefore ~300 m behind the source, the near-offset gap that no streamer geometry can close.

The core numbers

Streamers per vessel: 8-14 for production-class vessels. PGS/Polarcus/WesternGeco fleet averages around 10-12.
Streamer separation: 75-150 m cross-line. Narrower spacing improves near-offset azimuth but reduces total swath.
Streamer length: 3-8 km. 6 km is typical deep-water; 4 km for shallow-target surveys.
Group spacing: 12.5 m standard (each "group" is a short summed array of 4-16 hydrophones).
Vessel speed: ~5 knots = 2.5 m/s. Too fast stresses cables; too slow wastes ship time.
Shot interval: 25-50 m along sail line (10-20 s between shots), limited by listening time for the deepest target plus inter-shot decay.

What the geometry can and cannot image

A streamer survey delivers excellent inline imaging at near-vertical incidence but narrow azimuth (azimuths clustered ±10° from the sail line). Cross-dip structures (salt, fault systems oblique to the sail) under-image unless you shoot multiple azimuths (RAZ, §4.7) or use WAZ with multiple source vessels. The near-offset gap means AVO intercepts (§7.4 of the Processing book) are extrapolated, not measured, a limitation for gas-sand and CCS reservoirs.

Channel counts

10 streamers × 6 km × (one group / 12.5 m) = 4,800 live groups. Modern acquisition systems digitise every group every 2 ms, for every shot, an 8-hour survey day of ~2,900 shots with 6 s listening time produces tens of GB of raw data per day. Storage and processing bandwidth is enormous; this is where HPC (§5 of the Processing textbook) lives.

References

Sheriff, R. E., Geldart, L. P. (1995). Exploration Seismology (2nd ed.). Cambridge University Press.
Tenghamn, R., Brown, J. (2000). A new dual-sensor towed-streamer technology. SEG Annual Meeting Expanded Abstracts, 1-4.
Yilmaz, Ö. (2001). Seismic Data Analysis: Processing, Inversion, and Interpretation of Seismic Data (2 vols.). SEG Investigations in Geophysics 10.