Contact
For more information, please contact Christian Baresel.
christian.baresel@ivl.se
+46 (0) 8 598 564 06

Model study of pumping activities

Model study of pumping activities in the Baltic Sea (Hanöbukten and Gotlandsdjupet)

By employing a numerical 3D-model (AS3D) of the Baltic circulation, the impact on the stratification of down-pumping surface water in the vicinity of the location of the pump has been simulated and analyzed using oceanographic forcing data for the year 2004. In order to obtain a noticeable impact, a pumping capacity amounting to 25 m3/s was chosen. This value is about one order of magnitude higher than for the pilot version of the Wave Energized Baltic Aeration Pump (WEBAP), currently deployed in Hanöbukten in the southern Baltic.

An outspoken ambition has been to mimic the pumping conditions of the pilot version as closely as possible, exempting its 10-fold increased capacity and its slight positional offset to a location farther offshore than that which corresponds to the depth of the actual deployment site.

The interleaving of the plume - assumed in the model to be discharged at a depth of 80 m - has been modeled following the established theory augmented with the assumption that the down-pumped surface water is replaced by an incoming long surface wave towards the WEBAP platform. Conservation of volume demands that on the interleaving level an equal volume rate will flow out horizontally. This flow in the model is symmetrically super-imposed on the naturally occurring fluxes. Vertical entrainment of the rising buoyant plume in the depth interval between the discharge and the interleaving level will be the primary cause of the alteration of the stratification in the water column where the WEBAP is located. Since alteration and foremost lowering of the deeper density interfaces has been the major cause of biological concern, the unit deltat (i.e. deviation from the reference density of 1000 kg/m3) has been chosen as the basis for the evaluation.

The focus of the first part of this study has been the relative alteration of the stratification with and without a WEBAP stationed in Hanöbukten, and the results are graphically presented for a set of three representative but non-encompassing horizontal and vertical cross-sections in the near-field of the pump. These relative changes have been found to be modest, and only on one occasion (March 2004) was there a +1% relative aberration in deltat.

In order to perform a more encompassing comparison from a spatial point of view, the average density of the different layers from the surface to the bottom has been calculated in concentric horizontal slices departing from the site of the pump. A more frequent temporal basis for the analysis has been obtained by extending the analyzed periods to comprise the middle and the end of each month. This provides 24 time slices for which the average density differences with regard to depth of the model layers and radial distance to the pump have been calculated. These contour diagrams show that the minimum and maximum values of the absolute deltat-differences are restrained to the interval of –0.10 to +0.05 kg/m3. As the lowest surface water density corresponds to 5 kg/m3 (occurring during late spring), this means that the statistical average (over the entire near-field) relative aberration of deltat is well within a bracket of –2% to +1%.

An analogous analysis of an array consisting of 40 such WEBAPs — all with 25 m3/s pumping capacity — deployed in an equidistant array east of Gotland show that the impact on the density stratification in these open waters is small - in regard to differences to unaffected conditions less than 0.5%. After five years of pumping at the collective rate (1000 m3/s) for the array of pumps, the concentration of WEBAP-aerated water has increased to 1–3 % for a substantial part of the Baltic Proper intermediary depths.

With continued pumping under the same assumed conditions (constant pumping and only loss through water exiting the Baltic through the Danish straits), it will take about 35 years from the start of the pumping activities before the tracer concentration reaches a 1:5 dilution (corresponding to approx. 2 mLO2/L) that covers large parts of the bottom area on intermediary depth east of Gotland.

Model study of pumping activities in Stockholm middle center archipelago (Kanholmsfjärden)

By employing a numerical model based on discrete hydraulically coupled basins (CouBa) of the Stockholm archipelago, the impact of down-pumping surface water on the aeration and stratification of the basin where the pump was deployed has been simulated. This together with the spreading of the aerated water in the mid archipelago has been analyzed for the time period 1993 through 2004. A pumping capacity of 4 m3/s was chosen in accordance with the in situ pumping that has been going on intermittently since 2011 in Kanholmsfjärden. 

With the CouBa model approach combined with the implemented theory of plume interleaving, it is found that with water drawn from 5 m depth and discharged at 100 m, the yearly averaged interleaving level falls in the bracket of 50 to 70 m, i.e. as a temporal average above the measured position of the redox cline located at 70 m depth. However, the inherent model uncertainty can be appreciated to amount to 10 m and on occasion aerated water penetrates down to the bottom. The inherent model limitation of permitting persistent interleaving of numerous tiny layers has been mitigated by superposing two cases with the pumped layers merged both to adjacent layers upward and downward.

This result is not particularly sensitive to the scale length corresponding to the nozzle diameter of the vertical tube of the down-pumped aerated water (2.5 m) or a scale length (1.5 m) that also takes into consideration the horizontal screen at the end of the tube. This screen directs the out-flowing pumped water in a horizontal direction.

The results are presented as yearly averages of the 12 consecutive years, contour diagrams (time of year vs. depth with daily resolution) of the relative concentration of tracer-marked water that has passed the pump. This is not restrained to Kanholmsfjärden, but is also shown for a selection of basins distributed throughout the central mid part of the Stockholm archipelago. The minimal dilution of pumped water is obviously at Kanholmsfjärden, but Möja Söderfjärd displays a quite similar tracer concentration development. These two basins are hydraulically strongly coupled because of the broad and deep strait (90-m sill) connecting them. In all other basins in the archipelago presence of water that has passed the pump can be detected at a much higher dilution.

In these two coupled basins the minimal horizontally averaged dilution is found to be about 1% (or a dilution 1:100). In the vicinity of the discharge tube it can be inferred that the dilution is close to the primary dilution of the plume that varies between 10 and 20 times depending on the stratification. The interleaved modeled plume is instantaneously spread to the entire layer volume of its depth interval. In the real basin there is a gradient formed as the internal wave slowly progresses under constant vertical mixing out from the discharge point. The model thus exaggerates the horizontal dilution in the near-field by an order of magnitude. Since the pump is deployed at a deep plateau with 50-m isopleths nearby, it is likely that a sufficient oxygen supply could hypothetically tip the sediments from being anoxic if the redox cline should be higher than today’s 70-m depth.

The impact of the pumping activities on the temperature, salinity and density distribution in the other basins in the mid archipelago have been assessed by comparison with an analogous model run, with the only differing factor being that pumping was not performed. It is found that – as a yearly average – the alteration of the density stratification is mainly confined to the two major coupled basins. The pumping is also found to alter the net strait exchange flows in a small but systematic way.

Analysis of accessible measured temperature, salinity and oxygen concentration data cannot ascertain whether the bottom water in Kanholmsfjärden basin during the period 2012-05-16 and 2012-07-05 was oxygenated by inflowing bottom water from adjacent basins or by down-pumping of aerated surface water. Same data cannot, however, rule out that the WEBAP-pumping may have provided a beneficiary contribution to this end.

The website is administrated by IVL Swedish Environmental Research Insitute Ltd | © IVL