Collision between Norwegian frigate and tanker. Systems and mechanisms
At the beginning of the previous article (Collision between Norwegian frigate and Greek tanker) I noted that the investigation report is so detailed that it can be used to study the ship's systems. Let's take a look. I think the sea people will be interested.
First, let us recap which structures and organizations participated in the investigation:
NSIA: Norwegian Safety Investigation Authority, a government organisation for the investigation of accidents in all types of transport.
NDMA: Norwegian Defense Materiel Agency. You could call it the logistics department. Its job is to purchase military equipment, maintain it in good condition, and write it off when necessary.
NDMA Naval Systems Division: the division of NDMA specifically responsible for military fleet and its technical condition.
Defense Accident Investigation Board Norway: investigation of incidents in the troops.
Navantia: Spanish shipbuilding company specializing in both military and civilian shipbuilding. The fifth largest shipbuilding company in Europe. Builder of the Nansen-class frigate series.
Next, we need to somehow decide on the location of the compartments, which are mentioned every now and then in the report. Unfortunately, we were unable to find a drawing of the frigate with division into compartments. Or rather, such a drawing does exist, and it is very similar to the truth, but it was found not in official documents, but in the g-captain chat. The inscriptions there are on
There is also this drawing:
The ship seems to be the same, but the decoding of what each number means is not included with the drawing. Apparently, it's a secret.
Now let's briefly go over some of the ship's systems and devices that are mentioned in the report, and there I will also provide the conclusions of the technical examination of each system, if such conclusions exist.
Begin.
The report constantly refers to a certain IPMS.
Integrated Platform Management System - a multifunctional system that performs control and management functions on a ship, and at the same time registers and records everything in the world. Without any electronic systems, you just can't go anywhere.
The ship was built with a minimum possible crew of 120 people (the premises and rescue equipment are designed for 146) and has a high degree of automation. The crew uses IPMS to control and monitor almost any system on the ship, both during normal operation and in critical situations. The system records and remembers a huge amount of data - but with a period of 10 seconds, so some details can be missed during this intermediate period. Divers subsequently descended on the sunken frigate, who retrieved the memory blocks, and special specialists from a special institute restored almost all the data.
All IPMS records are collected in a separate appendix to the document, and some of them are reproduced here, but the appendix itself is marked as classified.
Electric power supply
The Fregat has 4 diesel generators with a capacity of 1000 kW each and two main distribution boards (MDB). Different pairs of DG and the corresponding MDB are located in different compartments.
The electrical installation is designed so that a failure of any equipment or consumer cannot lead to a blackout of the vessel – at least in theory. The main switchboards directly supply power only to large machines such as the bow thruster and local distribution boards called Load Centres (LC). LCs are distributed throughout the ship and supply consumers located nearby. All important consumers have dual power supply. The cable routes for such dual power supply are located as far apart as possible. Switching is done automatically or manually. All consumers can be managed via the IPMS system.
The two main switchboards can be connected to each other, or they can be independent. The Navy Administration, based on the 2015 incident, when a similar ship suffered a "blackout" when both main switchboards were working together, issued an addendum to the instructions that the main method of operation should be a separate method. However, at the time of the accident, both main switchboards of the frigate were connected, as in the picture.
Steering wheel control
You probably remember that after the collision the frigate had problems with its rudders. The commission stopped at this point.
The ship has two rudder blades, located behind the propellers and slightly offset from the shaft line (it is not said which one), and two independent steering machines. Each steering machine has two hydraulic pumps. In normal mode, one pump is sufficient to control the rudder, the second is in reserve. On this voyage, given the circumstances and the navigation area, all four steering pumps were in operation (this reduces the rudder shifting time by almost half). The pumps are started remotely via IMPS or in an emergency - from a local post.
The rudders can be controlled from four posts on the bridge, a separate joystick on the power plant control panel (PPC) in the engine room CPU, and in an emergency - from local posts in the steering compartment.
On the bridge there is a separate rudder control post (SSC) – this is the helmsman’s workplace.
In the first part there was a photo of this control panel taken by someone during the frigate's visit to Severomorsk, but at such an angle that the controls are almost invisible. Here you can see everything, but not very close, and the photo was taken after the frigate was lifted.
From this post you can control the rudders in Split Follow Up (i.e. separate operation of both rudders), Normal Follow Up (joint operation) or Non-Follow Up (NFU) mode.
Note. Follow Up: a mode where the rudder blade "follows" the control element, such as the steering wheel. For example, the helmsman turned the steering wheel 14.5 degrees to the right - and the rudder turned 14.5 degrees, and will remain in this position until the helmsman returns the steering wheel to "zero".
Non-Follow Up: for this mode, there is usually some other control mechanism - a handle with automatic return (tiller, that's what you can see in the photo), two buttons to the right and left, or something else with a similar action. The steering wheel moves while the corresponding direction button is pressed. Release the button, and the steering wheel remains where it was at that moment. To return it to zero, you need to press and hold another button.
If none of the above methods work, the steering wheel can be controlled from the emergency post in the steering compartment. There are also two methods for this: either from a similar remote control with buttons, in which case the bridge-tiller cables are excluded from the control chain, or completely manually activate the actuator, for example, by pressing the solenoid valve rod (fingers get tired very quickly). Conditions: there must be a trained person in the steering compartment, at least one steering pump for each rudder must be working, and there must be communication with the bridge.
The position of the rudders can be monitored on the multi-function display (MFD) at the helmsman's station, in the IPMS system and on separate indicators in different places in the wheelhouse.
There was also a separate rudder angle telegraph, which allowed commands to be given from the bridge to the steering compartment. The telegraph cables were laid on different sides.
At the time of the accident, the rudder was controlled from the SSC station on the bridge in Split FU mode, and all four pumps were operating.
After the collision, all four pumps stopped for 20 seconds (IPMS data), then only one pump, No. 2, started. After one minute and 13 seconds, three pumps were already working, except for No. 3. The steering gear then worked in this mode until 04:08, when Load Center 7 was de-energized. After that, only one pump worked for each steering gear.
The IPMS system recorded the movements of the rudder control joystick and the rudder response
Blue and yellow lines are the joystick movements, red and green are the position of the LB and PB rudders. As we can see, the rudders responded to the commands quite successfully.
Power point
It's not entirely clear story with water entering the main gearbox room and an unsuccessful attempt to stop the main engines from the bridge. Apparently, the commission didn't understand this either, so they devoted an entire section to describing the frigate's power plant.
Here we see what is called a CODAG type combined diesel/gas turbine plant, consisting of two diesels and one gas turbine. The propulsion is provided by two variable pitch propellers (VPP).
The IZAR BRAVO 12 diesels are four-stroke, 12-cylinder, V-shaped, with a power of 4500 kW each, manufactured under license on the basis of the Caterpillar 3612 engine and “specially adapted for installation on military ships” – whatever that means.
General Electric GE LM-2500 gas turbine with a capacity of 21,500 kW.
The main gearbox consisted of three main parts:
- a primary stage connected to two secondary stages and a gas turbine via a plug-in clutch;
- the secondary stage on the starboard side, connected to the primary stage, to the main engine of the PB and the propeller shaft with a variable pitch propeller;
- a similar secondary stage on the left side.
All this can operate in several modes, the details of which are classified information. But it is clear that the gas turbine is used when it is necessary to quickly reach full speed, which is 27 knots for the frigate, and the diesels, as the most economical part of the power plant, are used in cruising mode, that is, in order to obtain the greatest cruising range. It is possible that in the underwater target search mode, the frigate uses only one diesel engine, or even sticks out the bow thruster, which is marked on the drawing in the lower right corner as "retractable", and after that it becomes completely inaudible.
Usually the power plant is controlled via the IPMS system, i.e. remotely from the bridge or the engine room CPU. In case of a communication line break, the plant can be controlled from several local posts, the locations of which we will not list. In addition to controlling the diesels and turbine, there were local posts for controlling the propeller pitch.
The emergency stop can be initiated from several places, including the bridge and the control room. Such an event is noted by the IPMS recorder, however, nothing similar was found in the logs after the accident (see the photo of the IPMS screen after activating the emergency stop of a similar ship).
The commission then turned its attention to the design of the propeller shafts. The frigates built by the Spanish shipyard for various countries have similar technical solutions, but the frigates for Norway were somewhat different from the others. They had strict requirements for reducing their own noise and the ability to withstand the effects of underwater explosions. This entailed installing the main gearbox on a soft foundation and using flexible elastic couplings between the gearbox and the propeller shafts.
Now a little educational program. If a vessel has a variable pitch propeller, then in 99,999% of cases this means that its propeller shaft is hollow, and in this shaft a piston moves back and forth, which turns the propeller blades to the desired position. Such movement of the piston requires significant efforts, which are provided by hydraulics. Now we continue from the report.
The oil distributor, or OD-box (that's where the hydraulic oil for the CPP comes from), was placed in the intermediate shaft, located in the aft diesel generator room. This arrangement of the oil distributor differed from the Spanish F-100 frigates, where a similar device was placed on the bow side of the main gearbox.
From the OD-box, oil under pressure was directed through a two-layer pipe in the propeller shaft to a piston that changed the rotation of the blades, and through the same pipe it returned back to the oil distributor. This pipe changed its position together with the piston and was connected to a feedback sensor, which was located outside the propeller shaft.
The shipyard engineers also decided to install an intermediate hollow shaft between the OD box and the gearbox. The shaft had a diameter of 185 mm and passed from the aft DG room through the aft engine room to a flexible coupling in the gearbox room.
During the incident, it was noted that water was entering the main gearbox room through a flexible coupling. Investigation showed that water from the aft diesel generator room could have entered the main gearbox compartment through the hollow propeller shaft. In turn, it could have entered the hollow propeller shaft through the feedback sensor groove, which did not have any seals.
The fact that the OD-box oil distribution system could compromise the tightness of the frigate's compartments was not determined either during the design and construction of the frigate, or during the subsequent survey by the classification society DNV GL.
During the investigation it was discovered that in 2014-2015, the Helge Ingstad had experienced instances of steam leaking from the low-pressure compressor into the aft generator and aft engine rooms, causing the fire alarms in those compartments to go off. A smoke test was carried out, and the smoke leaked through the propeller shaft into adjacent compartments. This discovery was shared via email between the emergency response team, but was not reflected in the fault and discrepancy log.
Control of the rotation of the propeller blades
For this purpose, the frigate has two hydraulic stations located in the aft generator compartment. Each station has two main pumps, one auxiliary pump that maintains constant pressure, and one pump powered by compressed air (this is for emergency manual control). Here, too, a lot of interesting things happened.
Until 04:07, the pitch control was carried out from the central post on the frigate's bridge, after which it was switched to the Local position. At the same time, the corresponding switches on the local control post and the local control panel were not switched to manual mode.
Prior to the collision, the propulsion system was in cruise mode, providing a speed of approximately 17 knots. IPMS data shows the propulsion system operating mode before and after the collision.
Left VRS
After the blackout, both main gearbox oil pumps failed to start because both LCs feeding them were de-energized. When the gearbox oil pressure dropped, an emergency stop signal was sent to the LB main engine, and when it stopped, the CPP pitch automatically set to zero (blades in neutral position). At approximately 04:07, both pumps started automatically, and the CPP blades for some reason turned to -90% (i.e. almost full astern). The reason for this remains unclear.
Right VRS
After the collision, the starboard CPP control system lost contact with the IPMS, and remote control of the propeller pitch became impossible. The starboard CPP remained at +89% (almost full ahead). From 04:02:30, the frigate was moving forward at a speed of 5-5,5 knots, the main engine of the PB was running at low speed at 460 rpm. After grounding, the engine continued to run until 04:26, when it stopped. IPMS system did not record attempts to stop the engine.
At 04:05:59 the bridge control handles were moved from 65% to -18% for the right engine and 1% for the left engine. This had no effect because the main engine of the LB was not running and the connection between the IPMS and the right CP was broken.
Alternative mode of transportation
The frigate had two such options after the collision: a gas turbine engine, which in principle could have been started, and a bow thruster. As for the gas turbine, it was not working before the collision, and after the collision it received an automatic command for an emergency stop. The investigation did not find any technical reasons why the turbine could not have been started.
As for the NPU, it was quite officially a reserve means of movement. The documents do not indicate its power, nor the speed that the ship could develop with its help, nor the time required for its preparation. All this is classified information. But the principle is clear: the NPU extends from its shaft, receives power from the ship's diesel generators, and the ship is able to move.
communication
As you remember, not everything was well with her.
The frigate had the following communication systems:
- Audio unit (AU);
- Sound-powered telephone (SPT);
- Telephone;
- UHF;
- PA (Public address system).
The ASYM 3000A type audio unit (AU) was the frigate's primary means of internal and external communications. It is a digital system using some sort of "audio unit" in place. The report includes a photo of one of these units.
It was configured to create 12 internal "conferences", with local devices having different configurations. The AUs on the bridge and the CPU had access to all conferences. The system, oddly enough, had no backup power source and lost its configuration in the event of a power outage. After power was restored, all this had to be restored by pressing the Test/Lock button.
Note. I think I once encountered something similar on a small ship of the Wagenborg company. There was no PBX on the ship, and in the cabins and some rooms there were panels with a speaker, a button and a light. The speaker also served as a microphone. When you were called, the panel began to make nasty sounds like a frog croaking. I could be called from the bridge and from the CPU, respectively, and I could only contact them. To speak, you had to lean over the table, bring your lips close to the panel and hold the button down. Of course, there were no hassles with programming in this system. The impression was so-so.
Sound-powered telephone (SPT) – we call them battery-less paired telephones. To make a call, you need to turn the handle. Their advantage is that they do not require external power. On the frigate, this was the second most important communication system, duplicating the first, but it connected only the important control posts: the bridge-CPU-weapon-survivability control post - steering room.
Telephone. The ship had a telephone exchange that provided internal and external communications. In case of power failure, the telephone exchange was supplied from a UPS source, but provided only internal communications. To restore external communications (for example, to call headquarters), it took 4-5 minutes.
VHF radios were used primarily by emergency parties. VHF use was restricted in some areas of the ship.
PA (Public address system) – we call it loudspeaker communication. It is used to make announcements to the entire crew.
Durability and water resistance
This is a very important quality of any vessel, especially a warship. How was it with this on the frigate, and why did it sink so quickly? The commission was apparently very interested in this question, because a lot of attention was paid to the study of stability issues.
Note. The text uses the terms continuous damage and non-continuous damage, the meaning of which is not entirely clear to me. Perhaps these are the terms of the Norwegian Navy. I assume that non-continuous damage is damage that can be eliminated or minimized by crew members. For example, a fire can be extinguished, a patch can be applied to a hole, or the water supply can be restricted in other ways and pumped out.
The stability manual was originally written by Navantia according to the Royal Norwegian Navy Rules. Around 2014, the Navy suddenly decided to reclassify the frigate to DNV-GL class, so the NDMA unit of the Ministry of Defence had to rework the documentation in accordance with the DNV Rules.
For this purpose, they hired Polarkonsult AS, which provided DNV-GL with the required documents within the required timeframe, and in 2016 DNV-GL issued its approval of the stability calculations. At the same time, a decision was made to deviate from the intact stability requirement, according to which the GZ curve (in Russian, this would be “stability shoulder”) should have a range of at least 70 degrees. NSIA (the investigation committee) did not receive any explanation from NDMA as to why this requirement was cancelled, what consequences this had, or what compensatory measures were taken. However, after the incident, NSIA received calculations from Navantia, showing that the deviation had had a minor impact on the stability of the vessel.
The stability calculation is based on the rules (there is a long list of points and paragraphs). Nansen-class frigates have a waterline length of 121,4 meters, and according to the rules, calculations must be made based on a possible damage of 15% of the waterline, which for a frigate is 18,2 meters. In the worst case scenario, such damage would affect no more than three watertight compartments anywhere on the frigate's hull. More extensive damage would not necessarily lead to the ship sinking, but the "safety margin" required by the rules would not be met.
The ship was divided into 13 watertight compartments
The ship had stability documentation for all typical ship loading scenarios under normal conditions and in case of damage. This documentation had something called a "carpet plot". As far as I understood, it is some kind of analogue of our booklet on stability, but more visual. Its purpose is to help the crew in assessing buoyancy and stability in case of several damage scenarios. These are some diagrams in which you need to draw lines bordering the damaged area, and as a result you will get stability parameters for a given scenario. This is what the plot looks like.
The diagram shows that with the type of damage "continuous damage" (apparently meaning that this is damage that cannot be repaired) of three or fewer watertight compartments, stability is maintained in an "acceptable state", and in the middle part of the hull and near the bow of the ship "acceptable stability" is maintained with damage to four compartments. If the damage affects more compartments, the result will be "insufficient stability" or "the ship is lost". This plot did not provide any information about "non-continuous damage".
Quarterdeck (Q-deck)
Note. I don't know why Norwegians still use a term that comes from the sailing fleet, but apparently it's necessary. Basically, it's a section of the aft deck, slightly raised. On sailing frigates, the helmsman was located there, and from there the captain shouted "attack" or scolded the sailors. We call it the CP.
The spaces on this deck were not as watertight as expected and played a role in the sinking.
On Nansen-class frigates, the quarterdeck extends from frame 188 to 200 on the 2nd deck and forms part of compartment 13. From the quarterdeck, access is provided to the storeroom and several other spaces through hatches on the starboard and port sides.
On this deck there are six mooring line hatches and six working covers which are kept closed at sea. In addition, on the bulkhead of frame 188, on the sides, there are two spring-operated pressure relief valves. These valves are watertight in one direction only, from compartment 13 to compartment 12.
There is also a door called ATAS (Active Towed Array Sonar) with a hydraulic drive, controlled from a special control panel on the control panel. This door is open when the sonar antenna is released overboard.
In the original stability calculations made by Navantia at the design stage, the control deck was marked as watertight and weatherproof. Later, for some reason, the Ministry hired LMG Marin to review the original calculations and LMG reported that the ship did not meet the requirements of the Royal Navy Damage Stability Regulations because the control deck could not be considered watertight. At the same time, LMG relied on information provided by the Ministry that the control deck was not watertight due to the many doors and hatches on this deck. The Ministry thought about it and in 2004 (recall that the frigate was commissioned in 2009) reported to LMS that it had provided incorrect information and that all hatches and doors on the control deck were watertight. After that, LMG revised its calculations and recognized the control deck as watertight and the ship as compliant. It was this information that was later provided to DNV-GL when the ship was reclassified to its class.
According to the construction documentation, all penetrations (cable, piping, etc.) in the bulkheads of the command post deck were watertight. The same was stated for doors and hatches, but no documentation of any tests confirming this statement was provided.
The control deck could have made a significant contribution to keeping the ship afloat, but its watertight integrity had been compromised before the collision. It turned out that the ventilation valves on the control deck had been left in the open position, although they were marked Y (keep closed at sea).
The working hatches, the mooring line hatches and the sonar door were not marked at all. According to the crew, they were closed, but the commission has evidence from similar ships that there were problems with the tightness of these closures. Gaps appeared in the hatches after they were closed with battens, there were reports of damage to the covers and attempts to press them with hydraulic struts (jacks). There were problems with their maintenance, since due to the design features they had an outward slope.
Stability calculator
The calculator was created by the shipbuilder, Navantia, for all Nansen-class frigates as a tool for decision-making in case of damage. The software is embedded in IPMS. The calculator received data from level sensors in tanks ship, and information about damaged compartments was entered manually. The NSIA received information from the Navy that problems with the calculator arose both during the design stage and during operation.
The calculator has been viewed with mixed reviews on all ships. Crews have encountered issues with the complex user interface, inaccurate tank fluid levels, and problems interpreting regulations that needed to be resolved before the calculator could be put into service. The NDMA said that from the time the ship was commissioned until the November 2018 incident, neither the NDMA nor the Navy had given the calculator the attention it deserved in terms of operation, maintenance, training, and use.
In August 2017, three crew members of Helge Ingstad, who were taking a refresher course, were asked to evaluate a stability calculator and whether it could be used for the purposes for which it was intended. The response was as follows:
- Stability calculations are poorly described in the Norwegian Armed Forces regulations, manuals and publications. The information in some documents is outdated and needs to be revised.
- There are currently no training courses or courses available on how to use the Electronic Frigate Stability Calculator; therefore, the decision on how to do this is entirely up to each individual vessel. There are no courses or training courses available for crew on general stability calculations; therefore, competence on board is based on individual experience and educational level.
- Stability courses should be organized. The training should focus on the electronic frigate stability calculator, preferably with a set of user instructions. In addition, a unified approach to performing and organizing the calculations is necessary.
- The Stability Guide documents the stability of the Nansen class frigates in accordance with DNV GL requirements. The Guide is valid for a 5 year period between class tests. The Guide in its current form is very well suited for use in the case of "continuous damage" to several compartments, but is of little use for "non-continuous damage" cases.
- We were unable to test and confirm the stability calculator in the latest version of IPMS using the known loading conditions described in the manual. The reason for this is that there are too many bugs in the software itself. Therefore, we recommend using the calculator for training purposes only until the software has been debugged.
- The stability calculator "Helge Ingstad" has hardly been used due to insufficient training in the use of the software and insufficient knowledge about stability. Therefore, more attention should be paid to training. We also recommend some changes to the user interface to simplify the data entry process and make important information more visible.
Shortly before the accident, the authors of this memo sent a note to the responsible staff of the NDMA expressing concerns about the reliability of the stability calculator and the competence of the crew in its use. The team described this as a recurring and unresolved problem since 2006. In response, the NDMA stated that a solution to the problem was planned and would be solved in the near future, but did not provide an expected completion date. It was recommended that the crew contact the Naval Engineering and Safety Centre (KNMT NESC) or Navantia for assistance in training. As a result of the above circumstances, the stability calculator was not used either before or on the day of the accident. Following the incident, the NDMA requested Navantia to create new software.
Seawater system and drainage system
Here we are in for some wonderful discoveries.
These two essentially different systems are considered as one whole, since on the frigate they were closely connected to each other, and the drainage system could not function at all without pressure in the seawater system. That's how it is.
The system was designed based on three principles:
- Survivability: Components are designed to withstand various scenarios such as underwater explosions and extreme weather conditions.
- Redundancy: The system is divided into several sections, which allows maintaining significant performance even if one unit fails or is lost.
- Segregation: Different devices are located in separate watertight compartments and fire-hazardous areas to reduce the chance of damage to more than one device from the same accident.
The ship's designers solved the drainage problem in a very original way. The ship had a drainage system and a ballast system, but there were no drainage or ballast pumps. The ballast and water were pumped out of the rooms using powerful ejectors.
Note. Ejector pumps are found on any transport vessel and are usually used to drain holds, as they can suck out not only water, but also pieces of coal, wood, rags and other debris. What is it:
Pros: simplicity, no moving or rotating parts, no need for an electric motor with its quirks.
Cons: in the absence of working water it turns into a piece of metal, which we will see.
The system's performance is classified information, but the document contains a reference to the requirements of the Rules and Regulations for Surface Vessels of the Royal Norwegian Navy (RAR) and a calculation formula. According to the formula, the total system performance for a frigate should be no less than 340 cubic meters per hour.
The system was "combined" and included a "main" drainage system and a system for pumping out sludge and all kinds of contaminated water. All rooms with a sprinkler fire extinguishing system were equipped with a drainage system. It was also connected to the ballast system and the seawater system. Seawater was used to create a vacuum in the ejectors. The instructions from the shipyard stated: the main drainage system allows water to be removed from rooms below the damage control deck (see the figure above) and is capable of controlling the flow of water when extinguishing fires.
In total, the ship had six main ejectors and three independent systems of lower productivity, located in the helmsman's room, the vertical launch shaft compartment missiles and the placement of anchor and mooring winches.
Drawing of drainage system:
Almost all the valves in the bilge system were remotely controlled and had their own electric drive. These were: seven isolation valves between watertight compartments, six suction valves on the suction line in each engine room, six root valves after each ejector, and six driving water valves for feeding seawater to the ejectors. There were also regular valves with a manual drive, three in each compartment. They were painted black and were called black valves.
Water for "starting" the ejector (starting the ejector means creating a vacuum in it, which is necessary for pumping out water) came from the main seawater line.
The seawater system was designed as a ring line containing seawater at a constant pressure of 10 bar and having two loops, one on the port side and one on the starboard side. The loops could be connected to each other, but were usually isolated from each other by shut-off valves.
The pressure was maintained by six seawater pumps, one of which was diesel driven.
In the event of a failure, the affected section could be isolated from the rest of the system using remotely controlled valves. This required six Y-marked valves or three Z-marked valves to be closed, and at least two pumps to be running in the system, one for each loop. The design of the system was based on the assumption that the ship would be in state Y when at sea, and this was the case on the day of the incident.
Note. According to the Royal Navy Rules and Regulations, the letters X, Y, Z denoted the degree of protection of the ship. X – at the berth in peacetime, Y – at the berth in wartime and at sea in peacetime, Z – the highest degree of protection. Valves, doors, hatches, etc. were kept closed or open according to this condition.
The bilge and seawater system valves were normally controlled from the IPMS console in the CPU room, but could also be controlled from the local control station on Deck 2. The electrically operated valves could also be controlled manually in the event of a power failure. Many of the bilge system valves were located under the grating deck, the segments of which were bolted to the deck frame – that is, to access the valve, the grating had to be removed somehow (see previous figure).
In addition to the fixed drainage system, the ship had four portable electric pumps that required 440 V, 60 Hz. Each compartment had sockets for connecting these pumps, and according to the shipyard documentation, one socket could supply all four pumps via a splitter. The pump hoses could be connected to a DN4 water discharge pipe in each compartment on both sides.
Navantia also provided a maintenance and periodic testing program for the system and its components. Based on this program, NDMA prepared maintenance “work cards” according to which a “full” inspection of the system should be carried out every 5 years, and the remote valves should be tested every 6 months for their ability to close completely. The last inspection in 2018 did not reveal any non-conformities.
IPMS data for seawater system
Following the collision, the seawater system pressure dropped to zero. Isolation of the damaged area was complicated by the fact that remote control of several valves in the aft section of the ship had been lost. Before the seawater system was isolated, the IPMS operator started pumps 1, 2, 3 and 4, but the pressure in the system did not rise because water from the damaged system was flowing into the ship's compartments. The pressure at pump 4 was 10 bar, but valve MV-FM058 was closed and control was lost.
At approximately 0405 the damaged section between zones 2 and 3 was isolated by closing valves FM-MV047 and FM-MV165.
Valve 047 was reopened by the Damage Control console after approximately 20 seconds, causing the pressure in the system to drop again. The valve then opened and closed several times, causing pressure pulsations in the forward system, before finally closing at 04:07. The pressure in the forward system then stabilized at 10 bar. Navantia calculated that approximately 110 tonnes of water had entered through the damaged sections of the system.
IPMS data for ballast and bilge systems
Several bilge system valves lost communication with the IPMS and did not re-establish it after power was restored. These were the aft engine room isolation valve BD-MV046, the aft engine room ejector suction valve BD-MV049 and the aft generator room suction valve BD-MV056. They could not be controlled from either the IPMS console or the local console on Deck 2.
Between the second and third minutes after the collision, attempts were made from the propulsion control panel to activate ejector no. 1 (thruster compartment), no. 4 (main gearbox compartment) and no. 6 (stern generator compartment). The attempt failed because the damaged section of the seawater system had not yet been isolated. At approximately 04:05, an attempt was made from the auxiliary control panel to open valve 056 in the generator compartment, but this was not possible from any of the control panels.
Approximately six and a half minutes after the collision, control of valve BD-MV05, which isolates the compartment between the aft generator room and the aft engine room, was lost due to a power failure at switchboard LS7. At approximately 04:07, after the damaged area had been isolated, the seawater pressure for ejector no. 1 had risen to 10,2 bar, but the suction pressure ahead of the ejector was only -0,16 bar. An attempt was then made to use ejector no. 4 to pump water out of the group 3 ballast tanks by opening valve MV-BAL019 from the ACC station, but this too was unsuccessful as there was insufficient seawater pressure for the ejector to operate properly. The valve was soon closed.
At approximately 04:07, the isolation valves in the forward engine room and forward generator room were opened from the RCC console. The ejectors in these rooms had not built up sufficient suction pressure. The suction valve for the generator room ejector was closed to isolate the ejector from the bilge system, while the suction valves for the ejectors in the other rooms were open (see diagram).
At 04:08, the RCC console opened and closed the suction valve in the aft engine room for five seconds. At 04:14, the ACC console opened the suction valve in the thruster room, after which the suction pressure on the ejector dropped from -0.15 to -0.05. Twelve seconds later, the DCC console opened the isolation valve between the aft engine room and the main gearbox room.
At approximately 04:14, the ACC operator began using ejector no. 3 to pump 6,4 m3 from starboard ballast tank 4N02. This took 23 seconds. Navantia experts later calculated that this represented the total volume of water that had been pumped from the ship between the collision and the sinking (the detailed report was classified). The same operator then made an unsuccessful attempt to drain forward ballast tank 9L01 using ejector no. 1.
There was also insufficient suction pressure in the forward engine room, except in the forward auxiliary machinery compartment, where the ejector suction valve was closed. The ACC operator then opened the suction valve in that compartment at approximately 04:28, after which the ejector suction in that room dropped from -0,9 to -0,1 bar.
At approximately 04:38, 24 minutes after the thruster engine room suction valve had opened, the ACC operator closed it. This resulted in the ejector suction increasing from approx. -0,05 to -0,2 bar. The BDMV 015 isolation valve for the thruster was then closed and the ejector suction dropped again from -0,2 to -0,1 bar.
The ACC operator then closed the BDMV 025 isolation valve for the food waste collection system, whereupon the suction through the ejector in the forward auxiliary engine room increased from -0,2 to -0,7 bar. Shortly thereafter, the operator reopened the valve, whereupon the pressure in the ejector in the forward auxiliary engine room decreased to -0,2 bar. There is no indication that any further changes were made to the bilge system configuration.
Upon analysis of IPMS data for the main seawater system, as well as the ballast and bilge systems, Navantia concludes that no seawater was pumped out via the bilge system.
Imperfection of the ballast drainage system
The crews of the Nansen-class frigates have reported significant problems in the ballast drainage system, and classification society DNV GL has commented on them in connection with the upcoming periodic survey of the ships for class.
In 2014, in connection with the frigate’s reclassification to DNV-GL, six non-conformities were noted concerning the bilge system. The NDMA agreed that five of these needed to be addressed and a technical solution should be available by 2017. One of these was that, according to the DNV Rules, the bilge system should have a separate system for pumping out small quantities of contaminated water during normal operations and a high-capacity system for draining the engine room spaces. On the frigate, both systems were combined into one. It was found that the scope of the redesign was so great that the work was postponed until project funding could be secured and a design organisation established. These ideas were never put into action and the state of the system on the day of the accident was the same as when the ship received her DNV class.
Note. Further on, on several pages, there are discussions about the principles of interaction between various units of the Navy, relationships with the shipyard, DNV and various contract and subcontract companies, quotes from DNV Rules, SOLAS and Navy documents, inspection results, a description of the damage control training center and its programs... In general, I suggest skipping this. It is already clear that the drainage system did not work as expected.
But we will still give one quote:
Interviews with some of the Helge Ingstad crew revealed that prior to the accident, there was often too little time in practice to practice damage control scenarios in which several faults occurred simultaneously. The complex sailing program often made it difficult for the crew to stop the ship in the open sea and simulate a propulsion and steering failure in combination with other exercise elements. When performing damage control exercises, it was desirable to take the sailing program and the crew's need for rest into account. As a result, the exercise scenarios were often limited and adapted to these needs.
And finally, we come to an interesting section.
SPECIAL INVESTIGATION
Following the accident and the recovery of the vessel, a survey was carried out on board to establish the condition of the frigate at the time of its sinking and the state of its various systems. An extensive analysis of the IPMS data was also carried out, and some conclusions were drawn from this.
Note. The term complete shutdown is used here every now and then. I am used to understanding it as power outage, and therefore stopping any mechanisms. But most likely, in the document this word is understood as "turning off" the ship in a broader sense, for example, closing all doors, hatches, valves, ventilation openings, etc. So I will use the term "turn off the ship", no matter how strange it may sound. As you remember, before the evacuation, the ship's officers discussed the issue of complete shutdown and decided not to risk going down into the flooded rooms.
Stability calculation performed by the NSIA commission
NSIA has calculated the frigate's stability after the collision using the ShipShape program. The results are collected in Appendix D (it is not in the document, and I could not find it separately online). The calculations concern the time interval between the collision and the ship's grounding. The damage described in Section 2.2.1 of this document and in Appendix D was taken into account in the calculations. Damage caused by tugs was not taken into account, since the calculations show that the frigate would inevitably have sunk if the crew had abandoned it.
Main conclusions:
- failure to completely “turn off” the ship leads to its sinking;
- "turning off the ship" at the time of evacuation could have prevented the sinking;
- the ship's grounding on the rocks was not a decisive factor in its subsequent sinking, while the failure to "turn off" the ship after evacuation would have sunk the ship in any case;
- the flooding of the Q-deck had a significant negative effect on the stability of the ship, but was not a decisive factor in its sinking;
- the intermediate propeller shaft (with its ability to conduct water) had a negative effect on the stability of the ship, but was not a decisive factor in its sinking;
- the flow of water from tank to tank on opposite sides had a negative effect on stability, but was not a decisive factor in the flooding;
- If the frigate had not been held by the tugs, it would have started to drift. There is no indication that the frigate would have sunk any faster if it had not been held.
However, to prevent sinking, a "complete shutdown" of the ship was still required, which was not done:
The NSIA noted the following in its stability calculation:
- the lowest point of the breach was in the aft generator compartment (compartment 10) 260 mm below the waterline at the time of the incident. Damage to the side of the enlisted men's quarters (compartment 11) and the storeroom (compartment 12) also extended below the waterline. Presumably, compartment 12 flooded more slowly than compartment 11, but this does not change the main conclusions;
- at 04:07:40, one of the crew members, who was in the aft generator compartment, noticed that the hole was more or less at the water level. Calculations confirm this, as does the fact that the crew believed that they maintained control over the flow of water into the compartment until the moment the ship landed;
- calculations show that the ship's bow was subjected to a reactive force after it landed on the rocks, which resulted in the trim by the stern increasing. Calculations showed that at this point the lower edge of the hole was 100 mm below the waterline, which led to an increased flow of water into the stern generator compartment. This was also noticed by a crew member. The situation worsened, and the crew quickly lost control of the water flow. In turn, this led to flooding of the main gearbox compartment through the intermediate propeller shaft.
Checking maneuverability
Three maneuverability tests were conducted with two similar frigates, Roald Amundsen and Otto Sverdrup. The first test was conducted in calm waters in windless weather and was not documented, the second and third tests were conducted in conditions very similar to those on the day of the incident. I will not provide tables and figures, but the result is this: the frigate was able to maneuver after the collision, and up until 04:07:45, when turning to port, even with three of the four steering pumps operating, it had 5 minutes to avoid running aground.
Testing of the dehumidification system
In February-March 2019, the Helge Ingstad (i.e. after its raising) underwent a check of the drainage system valves to determine their ability to open/close. Two tests were also conducted to determine why the compartments were not draining effectively. The valve inspection and drainage system check were carried out by NDMA in the presence of NSIA representatives. Most likely, the valves were in the same condition at the time of the test as they were at the time of the evacuation.
All isolation valves were open except valve BD-MV015 between the forward auxiliary machinery compartment and the thruster room, valve BD-MV046 between the aft engine room and the main gearbox room, and valve BD-MV055 between the aft generator room and the aft main engine room. Several valves on the suction line located in flooded spaces were closed. For example, valve BD-MV056 in the aft generator room, BD-MV048 in the aft main engine room, and BD-MV032 in the forward main engine room.
The inspection showed that the overall performance of the dehumidification system was severely limited. It was found that three valves were not fully closed:
BD-MV010 – suction in the thruster room (the IPMS system received a false signal about the valve closing due to incorrect adjustment of the microswitch);
BD-V116 – manual suction valve in the food waste treatment room (was not closed);
BD-V027 – Manual suction valve in the pyrotechnic warehouse (valve seat defect).
As a result, the system was unable to create the necessary vacuum, which reduced the efficiency of drying.
Checking the performance of the dehumidification system
The test was conducted in January 2020 on board the frigate Thor Heyerdahl, whose dewatering system was similar to the Helge Ingstad. The objective of the test was to obtain data to compare the actual performance of the system with all six ejectors operating with the nameplate performance. The test was planned and conducted by the NDMA in cooperation with the Navy, with the participation of Navantia. The NSIA used Aker as a technical advisor. The results of the test have been declared classified information.
However, Aker made a conclusion that they did not or forgot to classify:
The observed pumping speed was too low for the purposes of the test and therefore did not comply with the specifications set for the vessel type. The deviations were sufficient to conclude that they could not be attributed to the accuracy of the measurements. The test also revealed deficiencies in that some valves could not be moved to the desired position or controlled remotely from the IPMS. This is a serious observation, as it indicates that the system could not be controlled properly. If it were not possible to close or open valves locally in a real situation, this could disable the system or significantly affect its operation. It was found that the vacuum and pressure readings of the working fluid in the ejectors in the IPMS and the readings from the local instruments did not agree with each other, so that it was not possible to determine with certainty whether the system was operating properly. The control system does not have any instrumentation to confirm the pumping speed.
In February 2021, NSIA received a response from Navantia regarding the test, which stated that the dehumidification system was functioning in accordance with regulations and requirements and that the test results were not “sufficiently representative” to draw conclusions about the actual performance of the system.
Checking the tightness of the Q-deck
In 2020 (i.e. after the raising), the frigate underwent a leakage test of doors, hatches, valves and, in general, all closures on the Q deck. Before the test, the doors were inspected, "serviced" and tested in action. The only thing mentioned about the test method is that it was a kind of "water test" using pressure corresponding to the depth of flooding of the stern. The test program and technology are given in an appendix, which is missing. The document only mentions the sonar antenna door, which showed a leak during a pressure test. However, a "standard test" using water through a fire hose showed no leaks.
NDMA Technical Investigation
The NDMA Naval Systems Division conducted a technical investigation into the incident. Most of the results of the investigation are classified, and there are no appendices to the report on this topic, but the main conclusions are available for review.
communication
The communications check focused on the bridge-engine room control room and bridge-tiller room communications between the collision and the grounding. With the exception of the loss of power, the following was noted:
Audio unit (AU): It is unlikely that the AU in the steering gear room was operational due to the break in the cable that was routed along the starboard side. It is also possible that the AU in the steering compartment lost power.
Sound-powered telephone (SPD): Based on our tests, we are unable to detect any defects or malfunctions that would likely indicate that the SPD telephone was inoperative after a collision.
Steering machine and rudder control
When power was restored to the 04SB main switchboard at 01:32:1, one of the LB steering gear pumps started automatically and the frigate was able to use the port rudder. After 04:02:22, three of the four pumps were operating, both rudders were operational and could be controlled from the bridge. An examination of the IPMS records revealed no indication that the selected Split FU control method was not operational. Due to the manner and location of the cabling, it is possible that the LSSSG001 – BRIDGE link was damaged or interrupted and the NFU control method for the starboard rudder was not operational. However, it cannot be concluded from the IPMS records that this method was selected for rudder control.
Steering wheel position indicator
It is highly probable that the PB rudder position indicators (three on the bridge and one in the steering engine room) were not working, including the display. No evidence was found regarding the LB indicator that it was also not working.
Steering control telegraph
The steering telegraph was most likely not working for the PB steering gear. There is no evidence found for the LB telegraph to indicate that it was not working either.
Multi-Function Displays (MFD)
The MFD in the helm station lost power and was not operational. The other displays were likely still operational.
Power point
PB propulsion plant: the RTU4112 (RTU — microprocessor unit for communication with the object, part of the IPMS system) immediately failed after the accident, as a result of which control of the starboard CPP via the IPMS system became impossible. Consequently, the propeller remained in the last known position of 89% forward. After raising the frigate, broken communication lines were found during its inspection, therefore, it was impossible to control the PB propulsion plant from the bridge using the joystick or the backup method. Since the feedback signal was also disrupted, it is impossible to determine whether the hydraulic pumps received 440 V power after the collision. The fluid coupling (FC) for the starboard engine was "opened" at 04:26:02 without a command from the IPMS.
Note. I once worked on a ship with two main engines, working on a propeller through a reduction gear. They were connected to the reduction gear using hydraulic couplings. The ship was an ice-going vessel, and hydraulic couplings were used when working in ice, since the impact of the propeller blade on the ice was somehow smoothed out by hydraulics and was not transmitted to the main engine. Some things remained in my memory, so I will say this:
In the description of events, there are two terms regarding the clutch: open and disengaged. Disengage cannot have any other interpretation than "disconnected, switched off". As for open, it is most likely meant that hydraulic oil is discharged from the clutch, without which, in fact, it cannot work. I believe that engage/disengage is a normal procedure when starting the power plant, and the "open" procedure is an emergency. Although it can be carried out by the operator's command, such a command should not be given with the main engine normally operating and connected to the gearbox. I remember that this sometimes happened to us when working in ice, and after such a shutdown it took some time until the clutch was filled with oil again and it could be engaged.
The most likely cause was a "slip" alarm (the difference in revolutions between the main engine and the gearbox) from the main engine control system, which was probably due to a sharp decrease in its revolutions. It also cannot be ruled out that the cause was water entering through the propeller shaft.
LB Propulsion Plant: The left engine FC clutch immediately disengaged after the collision. Technical experts suggested that the clutch disengagement could have been caused by a poor contact of the micro relay in the local control room, which opened during the impact and vibration that followed the contact of the vessels. The FC clutch also "opened", and the examination found no reason for this. It may have happened because both gearbox oil pumps stopped during the de-energization, when the load centres LC5/6 were disconnected. The pump with a mechanical drive from the gearbox also stopped working when the clutch "opened". Both pumps remained without power until 04.02.22.
The main engine of the LB received an emergency stop signal due to a drop in oil pressure in the second stage of the gearbox and remained in this state for the rest of the time.
The technical examination found no reason why the LB power plant could not have been started after the collision. No damage was found to the communication line "bridge-LB main engine". The reason why the FC clutch was in the "open" state was not found.
Control system of the VRS
Immediately after the collision, control of the starboard variator propeller from the bridge via the IPMS system was impossible, either in the normal or backup way. The only option left was manual emergency control from a local post in the aft generator room by direct action on the pitch change solenoid valves.
As for the left-hand CPP, no reasons preventing pitch control from the bridge were found until 04:06:21. It remains unclear whether emergency control from the local post was possible after that moment. Theoretically, it remained possible if the oil distributor was not flooded with sea water.
The -100% command to the LB screw could have come due to interference in the Profibus network (a network for controlling Siemens controllers, widely used in Europe for controlling industrial facilities). The influence of sea water penetrating the oil distributor cannot be ruled out either.
Note: Navantia concluded from its examination of the IPMS data that the probable cause of the 'full astern' pitch action was a short circuit in the cable that supplied the 'full astern' signal. As a result, when the backup control mode was automatically engaged after the collision, the control system received a command corresponding to the 'reverse' button at the local emergency control station being continuously pressed. However, the NSIA did not investigate this assumption as it was not critical to the outcome of the investigation.
Thruster (TH)
After the power outage, the IPMS recorded a command to stop the thruster. This signal remained active until the grounding. No physical reasons were found for the impossibility of starting the thruster: to turn off the emergency stop signal, it was necessary to manually restart the hydraulic pump. This was confirmed by a test on a similar ship.
After the blackout, both main switchboards are divided into 4 independent sections, and switches Q24/Q25 (main and backup power) for the control center are disconnected. After the accident, switch Q24 remained off until 04:08:23, i.e. the control center could not be used until that time. But since only one diesel generator was working, it was still impossible to use the control center due to insufficient power. The second generator was connected to the main switchboard at 04:13:51, when the ship was already sitting on the rocks. One explanation for such a late connection may be that the automatic switch of generator #2 had to be manually reset after the power outage. No technical limitations were found that would allow this to be done faster and then allow the control center to be used.
Drainage system and sea water system
The collision did not affect the ring main seawater line until it (it - a collision? Apparently, this refers to the development of a long hole in the side) spread to the aft generator room. Many small branches (from the main) were damaged, but this did not have a major effect. As for the aft generator room, the extent of the damage there would have made it very difficult to isolate the system. From a purely technical point of view, it would have been possible to move the point of isolation of the system further aft from the bulkhead between zones 2 and 3 at frame 90. This would have maintained pressure in the seawater system sufficient to operate the ejectors in the main gear room and the aft engine room.
IMF Internal Investigation
The Navy conducted its own investigation into the incident. It focused primarily on identifying nonconformities and their causes, with the goal of identifying systemic risk factors.
There is no information that this report is classified, but it could not be found. However, you can read excerpts from it.
Technical aspects and design
There are several significant inconsistencies related to the frigate's power system. Several orders for corrective action have been issued due to faults and defects in the system. Prior to the collision, the frigate was sailing with the main switchboards in combined mode, which is permitted by the design. However, the investigation showed that combined mode was a significant factor in causing the blackout after the collision. At the end of the third quarter of 2018, Helge Ingstad had 19 critical maintenance procedures outstanding. Five of these had expired.
Resources and Personnel
Here I asked Yandex for help, because the language became completely English-bureaucratic.
Some of the manning functions for the fleet are assigned to the ships themselves as vacancies arise. Combined with incomplete documentation of minimum manning requirements and safety competence, the responsibility for ensuring that ships are adequately manned is in practice assigned to the ship's commander.
The SAP tool is not intended to continuously monitor the collective competence situation on board ships.
Note. I spent a long time looking for what SAP is. The first part of the report also had a reference to SAP, and it listed the duties of the senior and watch officers and engineers. Therefore, we can conclude that SAP means Special Assessment Program - a gigantic document designed to assess risks - there is such a fashionable thing in modern management. On our ship we have something similar, invented by smart heads in the office. These are several volumes on the captain's shelf, which must be read upon arrival on the ship and signed. The larger the document and the more detailed it is, the easier it is to find the guilty parties later, if necessary.
The navy places fewer absolute demands on crew and competence than the complexity of modern ship operations requires. On-board operational and emergency safety relies heavily on careful and documented team training based on learning from Norwegian and allied experience, personal knowledge of colleagues and joint team training, and, to a lesser extent, documented individual competence. It is likely that ships occasionally employ personnel who do not have the competence to perform all the functions expected of them, and that important safety functions are, intentionally or unintentionally, performed by incompetent personnel. The risk is increased by routine practices in which personnel frequently change positions, particularly to fill vacancies as they arise.
And several pages are written in such language. In general, the SAP idea worked to the fullest extent - everyone is to blame. However, there is also a purely technical remark.
Radar
The switchboard that supplied the radars lost power, causing the X-band and S-band radars, the starboard rudder position indicators on the bridge and displays, and the navigation lights to stop working. The emergency procedures required for such an event were not followed.
The section then goes on to mention previous incidents involving warships: the grounding of the frigate Oslo in 1994, the fire on the minesweeper Orkla in 2002, injuries to crew members of a special purpose boat in 2010, the grounding of the patrol ship Ardenes in 2013. The same principle is evident: incident – investigation of the incident – issuance of recommendations for non-repetition. As, indeed, always and in everything.
I suggest we pause here. There are still the Analysis and Conclusion sections from Part 2 of the report ahead, then a separate Part 3, which is very short and probably doesn't contain anything special (I haven't read it yet), as well as a very short story about what happened to the ship and its commanders after everything. Stay tuned for more.
Information