Ecotoxicological Risk of World War Relic Munitions in the Sea after Low- and High-Order Blast-in-Place Operations

Abstract

Submerged munitions from World War I and II are threatening human activities in the oceans, including fisheries and shipping or the construction of pipelines and offshore facilities. To avoid unforeseen explosions, remotely controlled “blast-in-place” (BiP) operations are a common practice worldwide. However, after underwater BiP detonations, the toxic and carcinogenic energetic compounds (ECs) will not completely combust but rather distribute within the marine ecosphere. To shed light on this question, two comparable World War II mines in Denmark’s Sejerø Bay (Baltic Sea) were blown up by either low-order or high-order BiP operations by the Royal Danish Navy. Water and sediment samples were taken before and immediately after the respective BiP operation and analyzed for the presence of ECs with sensitive GC-MS/MS and LC-MS/MS technology. EC concentrations increased after high-order BiP detonations up to 353 ng/L and 175 μg/kg in water and sediment, respectively, while low-order BiP detonations resulted in EC water and sediment concentrations up to 1,000,000 ng/L (1 mg/L) and >10,000,000 μg/kg (>10 g/kg), respectively. Our studies provide unequivocal evidence that BiP operations in general lead to a significant increase of contamination of the marine environment and ecotoxicological risk with toxic ECs. Moreover, as compared to high-order BiP detonations, low-order BiP detonations resulted in a several 1000-fold higher burden on the marine environment.

Short abstract

Submerged munitions from World Wars are usually eliminated by remote-controlled “blast-in-place” (BiP) underwater operations. We show that BiP detonations lead to a contamination of the marine environment with unburned and toxic explosive chemicals.

1. Introduction

Millions of tons of munitions were dumped into the seas after the two World Wars.¹⁻³ For example, around 2 million metric tons of toxic conventional explosives (TNT and others) were sunk in the North and Baltic seas after World War II.^4,5 Until recently, the main threat of dumped war munitions was seen in unwanted and uncontrolled detonations, e.g., during offshore activities such as fishing, tourism, or the construction of wind farms and pipelines. However, a second important danger is now looming on the horizon. The munition shells are corroding, thereby leaking toxic energetic compounds (ECs) into the marine environment.^4,6−10 Hence, in addition to microplastics, heavy metals, pharmaceuticals, pesticides, and others, we have to deal now with a new and emerging group of pollutants within our oceans.

ECs like 2,4,6-trinitrotoluene (TNT), its conversion products (2-ADNT, 4-ADNT, DANT) and other nitroaromatics, are known for their toxicity and carcinogenicity. Various negative effects on a wide variety of aquatic organisms, such as sea urchins, shrimps, mussels, and fish, have already been demonstrated.¹¹⁻¹³ In humans, TNT can acutely cause jaundice and damage to the central nervous system, and in blood, it primarily reduces oxygen transport (conversion of hemoglobin to methemoglobin).¹⁴⁻¹⁸ Of greater relevance, however, is the fact that TNT and its conversion products are carcinogenic.^14,19 The MAK Commission (The Senate Commission for the Investigation of Health Hazards of Chemical Compounds in the Work Area) of the German Research Foundation (DFG) has defined category 2 (suspected human carcinogen) for the carcinogenic effect of TNT. It is feared today that ECs will accumulate in the marine food chain and directly endanger human health through the consumption of contaminated seafood.^20,21

The problem of environmental hazards is accentuated by the following fact. To protect today’s shipping traffic or the installation of pipelines and offshore facilities from uncontrolled and unforeseen explosions, remotely controlled demolitions (“blast-in-place” = BiP operations) of these dangerous World War relics are common practice worldwide. For this purpose, explosive charges are attached manually, which initiate the detonation of these munitions.

Underwater BiP detonations pose a serious threat to marine mammals such as harbor seals and harbor porpoises as high sound pressure and explosion-related shock waves can lead to severe injuries and hearing impairment in marine mammals in great distances from the detonation site.²²⁻²⁵

In addition, there are indications that upon underwater BiP operations these environmentally hazardous ECs may not completely combust but rather increasingly distribute within the marine ecosphere.²⁶⁻²⁸ By means of biomonitoring studies with blue mussels (Mytilus edulis) in the “Kolberger Heide”, a main munition dumping area in the Baltic Sea near the Kiel Bay (Germany), it was shown that BiP detonations with incomplete combustion lead to a 50-fold higher entry of ECs into marine biota, when compared to a pile of 70 corroding moored mines located in a distance of 500 m in the same area.^6,8,29

Following these problems, a conflict arises as to whether low- or high-order BiP operations should be carried out. On the one hand, low-order BiP detonations are performed with lower amounts of explosive charges to minimize the shock waves and to protect marine mammals. On the other hand, low-order BiP detonations, in contrast to high-order BiP detonations, are expected to leave higher amounts of noncombusted ECs (even smaller and bigger chunks) in the blasting area, which then dissolve in the water column and increasingly transfer into the marine ecosphere and biota.

The objective of this project was to investigate the pollution of the marine environment after low-order versus high-order BiP operations. For this purpose, two comparable mines in Denmark’s Sejerø Bay were blown up by either low-order or high-order BiP detonations by the Danish Royal Navy. These mines were remnants of some 400 ground mines that were dropped by the British Royal Air Force (RAF) aircraft to block the German shipping route from Aarhus to Norway during World War II (Royal Navy Historical Branch Portsmouth, Documents: “Gardening Tracings 1940–1945”). Since these mines pose a permanent threat to offshore activities such as shipping and fishing, they are being disposed by controlled BiP operations. To assess the level of contamination from BiP operations in general and to compare low- and high-order BiP detonations, water and sediment samples were taken by Royal Danish Navy divers before and immediately after the respective blast operations. The samples were then transferred to the toxicology laboratory and analyzed for the presence of ECs with sensitive GC-MS/MS and LC-MS/MS technology.³⁰ Our studies revealed that BiP operations in general lead to an increased contamination of the marine environment, which was dramatically high after low-order BiP detonations.

2. Materials and Methods

2.1. Historical Research

The current investigation is part of the European Interreg North Sea Region research project North Sea Wrecks (NSW) which focuses on sunken warships and their significance in environmental issues. That research project also includes investigations on environmental threats by corroding aircraft-launched mines dropped by the British military during WWII in Danish waters. Historical information was retrieved at the Federal Archives-Military Archives (BArch-MA) in Freiburg (Breisgau, Germany) (www.bundesarchiv.de) and from the Royal Navy and the Royal Air Force about mine laying in European waters during WWII.

2.2. Exploration of the Dumping Site

The mine field in the Sejerø Bay and across the Great Belt comprises about 400 ground mines (see more information in Section 3.1). Of these, around 14 are located in the Sejerø Bay offshore northwestern Zealand, close to the major shipping lane in the Great Belt (the T-route) and ferry routes from Jutland to Zealand (Figure 1A). Building on a close collaboration with the Royal Danish Navy during the NSW project, the opportunity arose to sample some of the mines in the Sejerø Bay in relation to a planned clearing operation by the Royal Danish Navy. Four mines were selected for further studies, the mines referred to as Charlie, Delta, Golf, and Mike. These four mines were of approximately the same type but differ with respect to corrosion status (Figure 2; Table 1). While the Charlie and Delta mines were reported by the Royal Danish Navy to be intact, underwater video and photo material from the Royal Danish Navy clearly show that the Golf and Mike mines were both open with the explosive material exposed directly to the seawater (Figure 2A,B).

Figure 1.

(A) Location of the four mines (Charlie, Delta, Golf, and Mike) in the Sejerø Bay in relation to major sailing and shipping routes (routes T and A, and ferry routes (black dashed lines and black arrows)) as well as the bathymetry.⁶⁶ The red box in the inset map shows the position of the study area centrally of Denmark. Green semitransparent polygons show protected Natura2000 zones, while the blue dashed line onshore Zealand indicates the southern limit of the UNESCO Geopark Odsherred that holds several protected landscape types. The gray arrow marks the dominant current direction through the Great Belt, while the red arrows show the sediment transport direction along the coast in the Sejerø Bay.⁶⁷ (B) Seafloor sediment type in the Sejerø Bay⁶⁸ showing a dominance of muddy sand and sand around the investigated mines.

Figure 2.

Photographs and seafloor scannings from the Royal Danish Navy showing the mines before and after BiP operations. (A, B) Underwater photographs of mines Golf and Mike before the BiP operations (June 2021) showing that both mines are open with explosive material exposed directly to the sea. (C, D) Underwater sonar scanning of mines Golf and Mike before the BiP detonations (June 2021). Note the flat and even seafloor surrounding both mines and some nearby stones at the Golf mine (C). (E) Pluton device for low-order detonation mounted at a mine. (F, G) Underwater photographs showing mines after low-order detonation (January 2022). Note the little damage to the mine casing (F) and the remaining explosive material scattered at and around the mine (G). (H, I) Photographs of high-order detonation (January 2022) showing first the water bulge rising to app. 5 m above the sea surface (H), followed by the sediment blasted to the air (I). (J) Underwater sonar scanning of the crater from the Mike high-order detonation (January 2022). The crater was 5.4 m in diameter and was approximately 3 m deep.

Table 1. Summary of the Mines Investigated^a.

name of mine	Charlie	Delta	Golf	Mike
type	ground mine MK I–IV	ground mine MK I–IV	ground mine MK VI	ground mine MK I–IV
position	56° 00.147′ N	55° 59.627′ N	55° 56.964′ N	55° 57.7182′ N
	011° 05.639′ E	011° 05.822′ E	011° 04.961′ E	011° 03.2198′ E
length	292 cm	292 cm	257 cm	292 cm
diameter	45 cm	45 cm	45 cm	45 cm
explosive type	amatol or minol	amatol or minol	amatol	amatol or minol
explosive amount	340.19 kg amatol	340.19 kg amatol	430.91 kg	340.19 kg amatol
	351.53 kg minol	351.53 kg minol		351.53 kg minol
state (open or intact)	intact (noncorroding)	intact (noncorroding)	open (corroding)	open (corroding)
BiP operation	none	none	low order	high order
burial	10 cm	10 cm	10 cm	5 cm
seabed conditions	fine sand	fine sand	fine sand with shells, shingle	fine sand with shells, shingle
water depth	22.5 m	20.8 m	19.5 m	18.5 m

^aPhotographs from mines Golf and Mike are shown in Figure 1A,B.

The Sejerø Bay comprises a generally sandy to muddy seafloor (Figure 1B), with water depths typically between 14 and 20 m and deepening in the Great Belt incision to around 40–50 m (Figure 1A). The investigated mines are therefore all situated above the mean wave base (except for ca. 20% of the time³¹) and hence mainly subjected to deterioration and degradation as a response to metal corrosion and ocean current erosion. The average temperature and salinity (in the years 2013–2016) of the water in the Sejerø Bay is reported to be respectively 9–10 °C and 19–32 psu.³² The mean current direction in the Sejerø Bay is toward the east and the south,³¹ while the average current velocity (in the years 2013–2016) is reported to be approximately 0.1–0.15 m/s.³² Higher current velocities, however, occur in relation to the strong water flows through the Great Belt, occasionally rising to around 0.6 m/s in the area of the investigated mines.³³

2.3. Details and Handling or Blasting of Mines

The four mines Charlie (C), Delta (D), Golf (G), and Mike (M) were visually examined by the Royal Danish Navy divers and then monitored with regard to the occurrence of ECs in the surrounding sediment and water. All four mines were ground mines located on a similar substrate (i.e., predominantly fine sand) and a preblasting burial of about 5–10 cm (Figures 1 and 2, Table 1). In January 2022, the two mines Golf and Mike were then low- or high-order blasted, respectively, by the Royal Danish Navy. Sediment and water samples from the blast sites were taken 30 min after the blast event (Section 2.4) and were later analyzed for ECs.

Mine Golf was low-order remediated by a pluton device with 250 g of TNT charge and a magnesium plate at 108 degree angle and 65 mm caliber. Observation of the sea bottom (Figure 2) as well as collection of sediment and water samples was performed approximately 30 min after the BiP operation by Danish Navy divers. Since approximately 20% of the energetic material remained unburned, the low-order BiP detonation was followed by a high-order BiP procedure to have the mine fully cleared.

The mine Mike was high-order blasted with 10 kg of TNT as the blasting charge. Observation of the sea bottom (Figure 2) as well as collection of sediment and water samples was performed approximately 30 min after the BiP operation by Danish Navy divers.

2.4. Sampling Strategy

Before blasting, in July 2021, sediment and water samples were taken by Royal Danish Navy divers at each of the four mines in three distances (0, 10, 50, or 0, 50, 100 cm) in four directions (front, back, left, right) as shown in Figure 3. Water samples at 0, 10, and 50 cm from mines Charlie and Delta as well as a few sediment samples from the Charlie mine were retrieved on 6 July 2021. Due to strong water currents and losing of bottles, it was however not possible to complete the full sampling program. Instead, sampling was completed on 27 July 2021, where the divers collected water and sediment samples at 0, 50, and 100 cm distances. The adjustment of the sampling distance was based on inputs from the divers who reported difficulties in differentiating between the 0 and 10 cm distances due to the strong bottom currents and low visibility in the area. After blasting of the Golf and Mike mines in January 2022, sediment and water samples were taken 30 min directly after the detonation in the direction “right” where the distances were extended up to 5 m (samples at 0, 1, and 5 m distances) (Figure 3).

Figure 3.

Sampling scheme for water and sediment samples before and after BiP operations.

All water samples were filled in 1 L polyethylene bottles, while the sediment samples were sampled in either zip-lock bags or polyethylene bottles. The zip-lock bags were only used on the initial sampling on July 6, 2021, but due to difficulties in handling the bags under water, zip-lock bags were replaced by plastic bottles for the remaining sampling. Sediment sampling was carried out by scraping the upper centimeters of the seafloor into the zip-lock bags or plastic bottles. All samples were kept cool on ice packs, then frozen on the same day at −20 °C in the laboratory at Aarhus University and later transferred to the Institute of Toxicology at Kiel University Medical School (Germany) for chemical analyses of TNT.

2.5. Chemical Laboratory Analyses

2.5.1. Materials and Chemicals

Trinitrotoluene (98.9%, 1 mg/mL, in acetonitrile/methanol (50:50)) was purchased from AccuStandard, New Haven. Isotopically labeled TNT (¹³C7, 99%; ¹⁵N3, 98%, 1 mg/mL in benzene, wetted with >33% H₂O) was purchased from Cambridge Isotope Laboratories, Inc., Andover. Acetonitrile (UHPLC grade, purity ≥99.97%), methanol (LC-MS grade, purity ≥99.97%), and water (LC-MS grade, filtered through 0.2 μm) were purchased from Th. Geyer (Renningen, Germany). Chromabond Easy solid-phase extraction columns 80 μm, 3 mL/200 mg, and 1 mL/30 mg (Macherey Nagel, Düren, Germany) were used.

2.5.2. Sediment Sample Preparation

Sediment samples were extracted for 100 g of wet sediment with the method as described in Bünning et al.³⁰ When the first measurement of a sample indicates an TNT concentration greater than 100 μg/kg, the unprocessed part of a sample was reprocessed using the method described in Bünning et al.³⁰ for 2 g of freeze-dried sediment. Eluates were filtered through 0.22 μm PTFE filters and concentrated to 1000 μL using a Christ RVC 2-25 CDplus rotary vacuum concentrator divided between two 1.5 mL amber glass vials as follows: 750 μL for GC-MS/MS analysis and 250 μL diluted with 750 μL of water (2.5 mM NH₄OAc) for LC-MS/MS analysis. Samples were stored at −20 °C.

2.5.3. Water Sample Preparation

Water samples were prepared by an adapted method according to the UDEMM Best Practice Guide.³⁴ One liter seawater was transferred into an EVA infusion bag (ICU Medical, Inc., San Clemente, CA) with 25 ng of ¹³C¹⁵N-TNT as an internal standard and allowed to flow through an unconditioned 3 mL Chromabond Easy SPE column, in the absence of light at 4 °C. The effluent was discarded. Columns were dried i.vac (0.5 h) and eluted with 4 mL of ACN. The eluate was concentrated to 1000 μL using a Christ RVC 2-25 CDplus rotary vacuum concentrator. Samples were divided between two 1.5 mL amber glass vials: 750 μL for GC-MS/MS analysis and 250 μL diluted with 750 μL of water (2.5 mM NH₄OAc) for LC-MS/MS analysis. Samples were stored at −20 °C.

2.5.4. GC-MS/MS Analysis

A TSQ 8000 EVO triple quadrupole mass spectrometer with an electron ionization source and a TRACE 1310 gas chromatograph (TG-5MS amine 15 m 0.25 mm × 0.25 μm column and quartz wool injection port liner) were used in secondary reaction monitoring (SRM) mode (Thermo Fisher Scientific Inc., Waltham, MA). Helium served as a carrier gas for the GC (1.5 mL/min) and Argon as a collision gas for the mass spectrometer. Samples (1 μL) were evaporated at 230 °C in splitless mode according to Bünning et al.³⁰ After 0.2 min at 100 °C, the oven was heated at 30 °C/min to 220 °C (0.3 min) and with 80 °C/min to 280 °C (1 min) for baking out the column. Spectra were recorded and analyzed in Chromeleon 7.2.10. TNT was detected at 3.41 min by the SRM transition m/z 210.0 > 164.1 (6 eV collision energy). The method-specific limits of detection (152 fg/μL) and quantification (502 fg/μL) were determined according to EUR 28099 EN.³⁰

2.5.5. LC-MS/MS Analysis

A Sciex QTrap 5500 triple quadrupole mass spectrometer with a Turbo V ESI ion source coupled to a UHPLC (Shimadzu Nexera LC-40D XS pump, Agilent G1316A column oven, CTC HTS PAL autosampler, VICI Cheminert 6-Port injection valve with 5 μL loop, RESTEK Raptor Biphenyl 1.8 μm column, 150 mm × 2.1 mm) was used. Spectra were recorded in Analyst 1.7.2 and analyzed in MultiQuant 3.0.3. Measurements start with a 5 min isocratic phase (40% H₂O (2.5 mM NH₄OAc)/60% MeOH) at 0.25 mL/min and 35 °C, then increased to 95% MeOH by 6 minute and maintained for 6 min. After the end of the measurement, the initial ratio is reestablished. TNT was detected after 9.04 min by the transition m/z 226.0 > 46.0 (−50 eV collision energy). Detection and quantification limits were determined to be 131 and 430 fg/μL.

3. Results

3.1. Historical Analysis

With the beginning of World War II, Germany used mine barriers as an important weapon to block the entrance to the Baltic Sea, via the Little Belt, the Great Belt, and the Øresund. Extensive barriers were laid in the North Sea as far as the Skagerrak. In order to ensure safe passage for warships and neutral merchant ships, gaps in the barriers were left in the mine field, and passing ships got order to use the ways (Figure 4). These barrier gaps were regularly checked by minesweepers, and thus the “Zwangswege” (“forced routes”) were created, which were marked by buoys, lighthouse-ships, and patrol boats.

Figure 4.

Historical map showing the “forced routes” through the Kattegat, the Danish straits, and the western Baltic Sea.⁶⁹

After Denmark and Norway were occupied by German troops on April 9, 1940, the forced route through the Great Belt from Kiel to Oslo was extended. It was given the name “Zwangsweg 28” (Figure 4). The port of Aarhus was significant due to its size and infrastructure and received the designation “Zwangsweg 28d”. To block the “Zwangsweg 28d” immediately after the occupation of Denmark and Norway by the Germans, Britain began an offensive mine laying operation that would last until the end of the war in 1945.

3.2. Description of the Mines that took part in the Trails of the Sejerø Bay

To prevent the shipping of military cargo from the German base in Aarhus further to Norway, at least 410 ground mines were dropped by RAF aircraft in the examined area of the Sejerø Bay, including the four designated mines Charlie, Delta, Golf, and Mike of the current study (Figure 1, Table 1). In most cases, these were mine types MK I–IV and MK VI, which share the same diameter of 45 cm but differ in length and explosive charge (Figure S1) (for details see Table 1). While the MK I–IV mine type has a length of 2.92 m and an explosive charge of 300–325 kg, depending on the type of explosive (charge 750 lbs of amatol or 775 lbs of minol), the MK VI is 2.57 m long and has an explosive charge of maximum 425 kg (charge: 950 lbs amatol) (Figure S1). The British military used amatol (a 40/60, 50/50, or 80/20 mix of TNT and ammonium nitrate) and minol (a 40/40/20 mix of TNT, ammonium nitrate, and aluminum powder). Those mixtures of TNT and ammonium nitrate were explosives of relatively poor quality but were used due to a shortage of TNT (http://www.navweaps.com/Weapons/WAMBR_Mines.php).

3.3. Energetic Compounds Found in Relation to the Investigated Mines

Table 2 summarizes the TNT water and sediment concentrations of all four mines and compares the enrichment factors after low-order and high-order BiP operations, while Figure 5 and Table S1 show the detailed results.

Table 2. Simplified Overview of Main Results.

name of mine	TNT water (ng/L)	TNT sediment (μg/kg)	blast	TNT water (ng/L)	TNT sediment (μg/kg)	x-fold water	x-fold sediment
	before blast (1 ma)			after blast (1 ma)
Charlie	0.4–16.3	0.8–3.7
Delta	0.5–23.0	0.1–47.0
Mike	1.1	0.7	HO	353	175	×320	×250
Golf	3.2	0.1	LO	1,000,000 (1 mg/L)	>10,000,000 (10 g/kg)	×312,000	×100,000,000

^aFor comparing prior and after BiP operation samples, values at 1 m distance were taken for mines Mike and Golf. For mines Charlie and Delta, values represent the average of all sampling positions.

Figure 5.

Detailed results from chemical analysis. (A) Mine Charlie before blasting, (B) mine Delta before blasting, (C) mine Golf before blasting, (D) mine Mike before blasting, (E) mine Golf after blasting, and (F) mine Mike after blasting. LOD = limit of detection.

3.3.1. Mine Charlie

Mine Charlie is a MK I–IV ground mine containing approximately 340 kg of amatol (or 351 kg of minol) with a calculated content of about 170 kg of TNT. The determination of leaking TNT in the surrounding water resulted in TNT water concentrations ranging between < LOD and 16.3 ng/L while TNT concentrations ranged from 0.8 μg/kg up to 3.7 μg/kg in the sediment (Figure 5A).

3.3.2. Mine Delta

Like Charlie, mine Delta is a MK I–IV ground mine containing approximately 340 kg of amatol (or 351 kg of minol) with a calculated content of about 170 kg of TNT. The determination of TNT in the surrounding water resulted in TNT water concentrations ranging between < LOD and 23.0 ng/L while TNT concentrations ranged from 0.05 μg/kg up to 1302.0 μg/kg in the sediment (Figure 5B).

3.3.3. Mine Golf

Mine Golf is a MK VI ground mine containing approximately 430 kg of amatol with a calculated content of about 215 kg of TNT. Before the BiP detonation, the determination of leaking TNT in the surrounding water resulted in TNT water concentrations ranging between 0.4 and >10,000.0 ng/L while TNT concentrations ranged from 0.1 μg/kg up to 35.0 μg/kg in the sediment (Figure 5C).

Mine Golf was low-order remediated with 250 g of TNT as the blasting charge, which resulted in a blunt sound and gentle bulge of water at the sea surface, apparently without sediment. There was no crater visible on the bottom floor and the mine metal caging were nearly intact (see Figure 2). TNT water concentrations increased several hundred 1000-fold up to 3,000,000 ng/L (3 mg/L) (in 5 m distance), while TNT concentrations in the sediment rose some million-fold beyond 10,000,000 μg/kg (10 g/kg) at all three measured distances (0, 1, and 5 m) (Figure 5E).

3.3.4. Mine Mike

Mine Mike is a MK I–IV ground mine containing approximately 340 kg of amatol (or 351 kg of minol) with a calculated content of about 170 kg of TNT. Before the BiP detonation, the determination of leaking TNT in the surrounding water resulted in TNT water concentrations ranging between 0.4 and 528.0 ng/L while TNT concentrations ranged from 0.1 μg/kg up to 524.0 μg/kg in the sediment (Figure 5D).

The mine Mike was high-order blasted with 10 kg of TNT as the blasting charge, which resulted, at the sea surface, in a loud sound and a sediment-laden water plume of approximately 5 m (Figure 2), and a crater on the sea bottom with a diameter of 5.4 and 3 m depth (Figure 2). TNT water concentrations increased some 350-fold up to 353 ng/L (in 1 m distance), while TNT concentrations in the sediment rose some 250-fold to 175 μg/kg (in 1 m distance; Figure 5F).

4. Discussion

Historically, the source of UXO was mine belts that were planted to defend the coastlines and shipping routes, shot-down fighter aircraft, leftover bombs that were dropped by aircraft to ensure a safe landing at home airports, training and direct combat, or ships that were sunken intentionally or unintentionally in combat.^13,4,35 In addition, disposal of the remaining munitions and chemical warfare materials in the North and Baltic seas was considered the best option following the World Wars.³⁶

Today, sea-dumped munitions should in principle be recovered and disposed of in an environmentally friendly manner using appropriate procedures. Some of the methods are still under development, including laser technology for welding the munition bodies or remote-controlled recovery systems that inactivate the munitions on site and incinerate them on a pontoon at the water surface. A major exception favoring blasting on site is an acute danger of self-detonation emanating from the UXO and the fact that the sensitivity continues to increase as the explosives age.³⁷ If this is not possible, it might be relocated under water to places where it does not perform any danger to shipping or other activities at sea and where it can be left on the seabed until future technical innovations for recovery have been developed.

Generally, there are several different ways for underwater BiP operations including deflagration, low-order BiP detonations, and high-order BiP detonations. These methods are defined according to their extent of conversion of the energetic material. However, underwater explosions are among the loudest point sources in the ocean, and UXO blasting is associated with a high level of noise pollution. In addition to sound waves, the explosion creates a shock wave that propagate through the water and the seafloor at supersonic speed which injure or even kill marine animals kilometers away.³⁸ The extremely rapid increase in pressure of the sound and shock waves and their amplitude are determining factors for the severity of injuries in marine mammals such as the harbor porpoise.²⁵

In principle, a bubble curtain could be used to reduce shock and sound waves and increase the safety area upon blast-in-place operations. The air bubbles reduce the shock wave and the extremely fast increase in pressure at the wavefront³⁹ such that the danger zone is reduced by up to 99%.⁴⁰ Disadvantages are the high costs and the enormous technical effort required.⁴¹

A second approach to reduce the sound and shock waves upon BiP operations is the application of deflagration methods or low-order detonations. Deflagration is a rapid burning of the explosives, where the rate of chemical reaction is considerably slower than in detonations, with a maximal reaction velocity of less than 1000 m/s.⁴² While upon deflagration no shock wave is generated, the extent of contamination should be high because unreacted or only partially reacted explosives are released into the environment.

As an alternative to deflagration, there are deactivation methods in which the ignition chain is interrupted. The detonator is separated from the ammunition by lasers or small explosive charges, which enables subsequent handling of the UXO. The latter defusing technique is also a low-order method where the fuze is blast off the main charge by a limited detonation.⁴³

A low-order BiP detonation is a procedure against the main charge with the aim of disarming (separation of the detonation chain) or destroying the explosive device. In this case, part of the main charge is also reinvolved, but this should not detonate. As a result, the reaction speed is higher than upon deflagration and a weak shock wave is produced. In addition, at low-order BiP detonation, unreacted or only partially reacted explosives are released into the environment.⁴⁴

A high-order detonation is characterized by a high detonation velocity and a shock-wave-like, uniform dispersion of all combustion products with extensive conversion of the explosive. In explosive ordnance disposal, this refers to the targeted detonation of the explosive ordnance by triggering a detonation with the aim of completely detonating the main charge. However, the potential chemical contamination is a general environmental risk with all blasting of old munitions.⁴⁴

Due to the ongoing corrosion of the metal shells, all four mines were shown to leak explosive chemicals into the environment already before the BiP operations. EC concentrations around all four mines varied on average in water between 0.4 up and 23.0 ng/L and in sediment between 0.8 and 47.0 μg/kg. In two water samples of mine Golf and one sediment sample of mine Delta (directly at 0 cm), higher EC concentrations were observed, probably due to a contamination with solid EC fragments.

EC concentrations increased after both blast operations. Blasting of mine Mike by high-order BiP detonation resulted in EC water concentrations up to 353 ng/L and in EC sediment concentrations up to 175 μg/kg over a 1 m distance on the right-hand side (Figure 5F). Compared to the situation before blasting (at 1 m on the right-hand side of the mine Mike; Figure 5D), the measured concentrations increased by factors of 350- and 250-fold, respectively (Table 2). The same extent of ECs was also found on the right side of the mine at a distance of 5 m in the water (88 ng/L) and sediment (260 μg/kg), suggesting an even distribution of the energetic chemicals in the environment.

Blasting of mine Golf by low-order BiP detonation resulted in EC water concentrations up to 1,000,000 ng/L (1 mg/L) and in EC sediment concentrations up to >10,000,000 μg/kg (10 g/kg) in a 1 m distance on the right-hand side (Figure 5E). Compared to the situation before the blasting (at 1 m on the right-hand side of the mine Golf) (Figure 5C), the release of energetic compounds increased by a factor of 315,000 and 1,000,000, respectively. Moreover, comparable high concentrations of ECs were found on the right side of the mine at a distance of 5 m in the water (3,300,000 ng/L) (3.3 mg/L) and sediment (10,000,000 μg/kg) (10 g/kg). These findings suggest an even and strong contamination of the marine environment at even greater distances (i.e., >5 m). Whatsoever, while the high water concentrations directly result from dissolved ECs, the high concentrations in the sediment may also result from EC fragments that are present in the sample and released ECs upon the sediment extraction procedure.

To discuss the risk associated with the presence of ECs in aquatic environments, both the concentration of ECs in the environment and information on toxicity thresholds in biota are necessary. Toxicity data on freshwater and marine species end-points are usually derived from field studies or laboratory experiments; the latter with TNT contaminant-spiked water or sediment in concentrations usually significantly higher than those expected in the environment. This implies that the resulting effects represent acute rather than chronic adverse effects.^{11,12,45−48}

In these studies, TNT caused decreased population growth of cyanobacteria and green microalgae, at concentrations ranging from 0.75 to 18 mg/L.⁴⁸ Numerous aquatic toxicity studies have reported that TNT caused decreased survival and offspring production in aquatic invertebrates with toxicity ranges between 1 and 6 mg/L, for example, in sea urchins and mysid shrimps,⁴⁹ bivalves,^9,45,50−52 copepods,⁵²⁻⁵⁷ or polychaetes.⁴⁹ The marine flatworm Macrostomum lignano was shown to stop feeding and starve when exposed to TNT at 3 mg/L.⁵⁸

An interesting observation was made by Strehse et al.⁹ and Jacobsen et al.⁵⁹ While Strehse et al.⁹ found a concentration-dependent upregulation of the carbonyl reductase gene in blue mussels (Mytilus spp.) at TNT concentrations of 0.31 mg/L and above, Jacobsen et al.⁵⁹ proofed an upregulation of the carbonyl reductase gene as a response to TNT exposure in Daphnia magna at concentrations starting at 0.5 mg/L. Carbonyl reductases are members of the short-chain dehydrogenase/reductase (SDR) superfamily and provide a defense mechanism against “carbonyl” stress and reactive oxygen species (ROS). TNT is a known source of ROS, which causes oxidative stress in aquatic ecosystems. Thus, carbonyl reductases may constitute a defense mechanism against ROS products induced by TNT.

Of importance is the fact that the carbonyl reductase gene was induced in blue mussels in a 58-day field biomonitoring study at a major dumping site in the Kolberger Heide in the Baltic Sea (near Kiel, Germany). The authors concluded that the gene coding for the carbonyl reductase enzyme may serve as a molecular biomarker in marine biota to evaluate TNT contaminations in field studies.⁹

Adverse biological effects and lethal toxicity of explosive chemicals (in the low mg/L range) have also been reported for freshwater and marine fish.^{46,49,60−62} A decreased survival of larval freshwater, marine, and estuarine fish was observed at concentrations ranging from 0.4 to 28 mg/L. Mariussen et al.⁶³ investigated the effects of TNT in juvenile Atlantic salmon (Salmo salar). Mortality, severe hemorrhages in the dorsal muscle tissue near the spine, and effects on blood parameters such as glucose, urea, hematocrit, and hemoglobin were observed in fish exposed to TNT water concentrations of 1 mg/L. Competitive inhibition of 7-ethoxyresorufin-O-deethylase and 7-methoxyresorufin-O-deethylase by TNT in vitro was demonstrated for livers from three flatfish species (Limanda limanda, Pleuronectes platessa, and Platichthys flesus) sampled from the Baltic Sea.⁶⁴

The mutagenicity and carcinogenicity of nitroaromatic compounds have also been shown in laboratory studies with zebrafish embryos (Danio rerio).⁶⁵ Exposure to TNT as well as 2- or 4-ADNT revealed a high proportion of chorda deformations at concentrations in the milligram per liter range in the surviving embryos. In addition, genotoxicity of the nitroaromatic compounds was shown by a comet assay, with cells isolated from whole embryos after 48 h of exposure in vivo. Significant genotoxicity was induced by all three tested compounds at 0.1 and 1.0 mg/L.

Relatively few studies addressed the toxicity of sediment-associated explosives to freshwater and marine invertebrates and to fish in exposure to spiked sediments. The toxicity of sediment spiked with TNT to marine polychaete, estuarine amphipod, freshwater midge, amphipod, oligochaeta, and estuarine fish occurred within a wide range of concentrations (37–508 mg/kg).⁴⁸

The concentrations of TNT in water and sediment measured in unaffected munition dumping areas so far are at least several 1000-fold lower than effect levels reported in marine species in laboratory studies. However, according to our findings, it should be made clear here that any BiP operation will result in a significant increase in TNT water and sediment concentrations as well as elevated body burdens in marine biota that may be fatal to all marine life in the vicinity of the BiP operation sites.^8,29

Clearly, the measured concentrations after low-order and high-order BiP detonations will be mixed during and after detonation and will dilute over time. Beck et al.⁷ have shown that dissolution fluxes show large variability, ranging from 1 to 3 mg cm^–2 day^–1 and conclude that the surface of exposed underwater munitions will retreat at a low rate on the order of 1–5 mm per year. However, the concentrations measured in our study after the BiP operations dramatically increase in both water and sediment to values known as effect levels reported in marine species in laboratory studies. Consequently, as discussed above, it should be made clear here that this increase in TNT water and sediment concentrations will lead to elevated body burdens in marine biota and may be fatal to all marine life in the near vicinity of the BiP operation sites.^8,29 It is also clear that low-order BiP detonations with a million-fold increase in TNT concentrations (in the milligram per liter range in water and double-digit gram range per kilogram of sediment) can cause not only chronic but also acute health damage in the marine biota.

Strehse et al.,⁸⁸ Appel et al.,⁶ and Maser and Strehse²⁹ performed biomonitoring studies by exposing blue mussels (Mytilus spp.) in the munition dumping site Kolberger Heide near the Kiel Bay (Baltic Sea, Germany). While mussels directly exposed at corroding moored mines contained ECs in the 4–6 ng/g tissue range,⁶ mussels exposed at free-lying pieces of hexanite (German: Schießwolle; derived from BiP operations and submerged hexanite chunks) contained ECs in the 350 ng/g tissue range.^8,29 Obviously, the corroding metal shells of munitions items reduced the dissolution and distribution of ECs within the marine environment, whereas free-lying TNT led to a 50-fold higher burden of adjacent biota. This was the first report which recommended “do not blast” as a strategy to protect the marine environment from human activities 70 years after World War II and provided the idea for the present experimental setup.

While our studies provide unequivocal evidence that BiP operations, especially low-order BiP detonations, result in an incredibly high EC pollution burden of the marine environment, further studies are necessary to measure the distribution of ECs in wider distances from the blast site and taking into account that dilution and mixing of ECs take place continuously. To recover submerged World War munitions in an environmentally friendly way, there are already alternatives to BiP operations available or are under technological development, including remote-controlled laser systems to open the metal casings on site, followed by recovery and subsequent burning on the water surface, e.g., on pontoon ships.

Glossary

Abbreviations

BiP

blast in place

ECs

energetic compounds

TNT

2,4,6-trinitrotoluene

Data Availability Statement

Data are contained within the article.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.est.3c04873.

• Black and white photographs and cross-section drawings of the investigated two British Ground Mine types MK I–IV and MK VI (Figure S1) and the results from the chemical analyses (Table S1) (PDF)

Author Contributions

Conceptualization, E.M., K.J.A., and J.S.S.; methodology, E.M., K.J.A., J.S.S., O.R.C, and T.H.B.; validation, J.S.S. and T.H.B.; formal analysis, T.H.B. and U.W.; investigation, E.M., U.W., J.S.S., and T.H.B.; resources, E.M., J.S.S., K.J.A., and O.R.C; writing—original draft preparation, E.M.; writing—review and editing, E.M., K.J.A., and J.S.S.; visualization, K.J.A., U.W., J.S.S., O.R.C, and T.H.B.; supervision, E.M. and K.J.A.; project administration, J.S.S. and E.M.; and funding acquisition, E.M., K.J.A., and J.S.S. All authors have read and agreed to the published version of the manuscript.

The authors declare no competing financial interest.