the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Entrainment of oil droplets by random waves
Abstract. A series of laboratory experiments was conducted to investigate the oil droplet size distribution generated by the entrainment of oil under random breaking waves. Experiments were performed using three different oil types and two distances between oil release and droplet sampling, to examine the evolution of droplet size distributions with prolonged wave action. The random wave field was designed to replicate open-ocean conditions through a combination of a JONSWAP spectrum and superimposed sine-wave representing swell.
The temporal evolution of the volume-based median droplet diameter, d50v, for the naphthenic Troll B oil indicated a pronounced dependence on the distance from release to sampling point, with characteristic droplet size reduced by about a factor 2 when subject to wave action for a longer distance. In contrast, the results for Oseberg Blend, with a higher wax content, showed substantially weaker dependence on release location, the characteristic droplet size differing by only a factor of 1.3 between the two release sections.
The predictive performance of several existing droplet size models was systematically evaluated against the experimental data. The relative prediction errors span a wide range, from as low as -2 % to over 8000 %, highlighting significant uncertainties in current modeling approaches. Sensitivity analysis demonstrates that model predictions are strongly influenced by the choice of characteristic length scale, and the choice of significant wave height used to derive the velocity scale. A comparison between time-resolved predictions and experimentally measured maximum droplet sizes indicates that rare, high-energy events dominate the maximum droplet diameter, highlighting fundamental limitations of deterministic breakup models under random wave conditions.
- Preprint
(40660 KB) - Metadata XML
-
Supplement
(1093 KB) - BibTeX
- EndNote
Status: open (extended)
-
RC1: 'Comment on egusphere-2026-676', Anonymous Referee #1, 07 Jun 2026
reply
-
AC1: 'Reply on RC1', Tor Nordam, 18 Nov 2026
reply
We thank the Reviewer for the positive assessment of our manuscript and for the constructive comments. As the discussion is still open, we outline here the revisions and modifications we plan to implement; these will be added in the revised manuscript together with any changes and revisions arising from the second review.
Comment 1
We also agree that the oil slick thickness measurements in time and space are practically difficult, as we mentioned in the Introduction section, lines 71-73 of the original manuscript.
We have demonstrated the quantitative sensitivity analysis in Figure 9 for hs (oil slick thickness) in the range of 0.08 to 3.125 mm by comparing the models performance (see Figure 9 and lines 350-360 of the original manuscript). However, to specifically address the uncertainty associated with the models performance, we will add a short paragraph and present a table to Section 4.2. For the three models that use hs as the characteristic length scale, the predicted dv50 scales as hs0.40 for the Johansen model (Johansen et al., 2015), hs0.34 for the refitted Johansen model, and hs0.12 for the Mostaani model (Mostaani et al., 2026b), independent of oil properties over the range of hs considered in Figure 9 (0.08–3.125 mm). These scaling relations can be directly determined from Table 1 of the original manuscript.
Hence, the predicted dv50 from the Johansen model, the refitted Johansen model, and the Mostaani model varies by factors of about 4.33, 3.48, and 1.55, respectively. Higher factors for the Johansen model and the refitted Johansen model compared to the Mostaani model are consistent with our statement in lines 354-355 of the original manuscript:
"As shown, the predictions of both the original and refitted Johansen et al. (2015) models exhibit substantially higher sensitivity to oil slick thickness compared to the model of Mostaani et al. (2026b)."
Notably, the reported relative errors for these three models in Table 6 of the original manuscript are obtained by assuming an oil slick thickness of hs = 0.20 mm, whereas the predictions of Li et al. (2017b) and Wang et al. (2024), which use dRT as the length scale, are independent of hs.
Comment 2
We thank the Reviewer for this suggestion and will add a discussion subsection called "Modelling implications" addressing the two points:
(a) Approach to equilibrium. The Section B measurements for Troll B show a more stable dv50, a narrower distribution and a higher droplet number, consistent with further development of the droplet size distribution. However, with two release distances, with estimated exposure times of approximately 8 s and 37 s, our data cannot establish whether the distribution at Section B represents an equilibrium state. We will discuss this limitation, the role of buoyant resurfacing, and re-entrainment at a fixed sampling depth. The point the reviewer raises about equilibrium timescales is also worth discussing further.
We will also compare our results with recent in situ measurements during an experimental oil spill in the North Sea (Nordam et al., 2026), where a qualitatively similar transition was observed, from large droplets under the thick leading edge of the slick to smaller, more uniform droplet sizes further into the slick.
(b) Operational implications. We will add a brief discussion of the implications of our results for operational oil spill models, which commonly apply an equilibrium droplet size parameterisation at each breaking event. In particular, we will discuss the sensitivity of such parameterisations to the choice of wave height and slick thickness, and the observation that, for identical oil and wave conditions, the measured droplet size distribution depended on the distance from the oil release, which is not represented in equilibrium models.
Comment 3We thank the Reviewer for this suggestion. We will improve the readability of Figures 6 and 7 by increasing the resolution of the heatmaps, adjusting the color scale range of the heatmaps, and enhancing the contrast of the dv50 (t) and dv50 [All] curves relative to the background..
Citation: https://doi.org/10.5194/egusphere-2026-676-AC1
-
AC1: 'Reply on RC1', Tor Nordam, 18 Nov 2026
reply
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 817 | 481 | 80 | 1,378 | 224 | 78 | 161 |
- HTML: 817
- PDF: 481
- XML: 80
- Total: 1,378
- Supplement: 224
- BibTeX: 78
- EndNote: 161
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This manuscript presents a laboratory investigation of oil droplet entrainment under random breaking waves using three weathered crude oils and repeated flume experiments. The authors provide a valuable dataset of droplet size distributions measured under irregular wave conditions and evaluate the performance of several existing droplet-size prediction models (i.e., d50 and dmax).
The study is well designed, the experimental procedures are clearly documented, and the results provide useful insight into the evolution of droplet size distributions under prolonged wave action. In particular, the comparison between different release locations provides an interesting perspective on droplet equilibration processes and repeated entrainment. The manuscript is generally well written and fits well within the scope of Ocean Science.
I recommend publication after minor revisions.
Overall, I find the manuscript scientifically sound and suitable for publication after minor revision.