Change search
ReferencesLink to record
Permanent link

Direct link
Brand i kabelkulvert: En analys av dimensionerande bränder i en kabelkulvert
2013 (Swedish)Independent thesis Basic level (professional degree), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

At Akzo Nobel Stockvik there is a cable-culvert where many cables run. There is no oversight of the culvert and no impact-analysis in case of a fire. Electricity is fed through the culvert to AGA-Gas. Penetrations made from the culvert to other areas are not documented and the risk of the spread of smoke and heat to other rooms is possible. The report aims to assess the impact of a fire in the culvert and perform calculations on various fire scenarios in the culvert. The results from these should be used as an input-file in the computer program Fire Dynamics Simulator (FDS).The work has been limited to the culvert and related areas. The FDS simulations are used solely to calculate the spread of smoke and temperatures.The work has been performed in different steps by a predetermined method. This was developed through discussions between the supervisor, Björn Sundström, and writer, Anton Hörnqvist. The flow consists of an object visit with photographing and measuring of the culvert. Thereafter, subsequent literature studies, calculations and assumptions. The calculations were used as input file in FDS.Literature studies focuses mainly on statistics on fires caused by electricity and cables. This includes different causes of errors leading to fires in cables.The results of the visual examination was made by a 3-D model. Since there are no drawings of this culvert the model has been used as the orientation of the culvert. The model has later been divided into sections for easier reference to the specific area. The inspection also revealed a number of shortcomings regarding safety. These have been summarized in a checklist.An impact assessment was performed by using personnel in the field. This showed that a number of areas would be without electric power supply in case a fire should occur in the cable culvert. The most severe consequences would be power loss for AGA gas and the SKW:s (Stahl-Metallurgie Holding AG) finishing operations.Calculations were made to obtain a heat release rate curve. Quickly it was found that due to the windows along with the limited amount of oxygen it would be likely that the fire would become ventilation controlled. Therefore two heat release rate curves were calculated. One scenario was that the fire would continue until all the oxygen in the culvert was used, and then burn itself off. The second scenario was that the windows bursts and therefore leads to the continuation of the fire, but with the limited amount of oxygen supplied through the windows. In Scenario 1 the maximum power of 2.8 MW is developed. In scenario two the effect of the fire decreases, when the fire get ventilation controlled, down to 1.8 MW.The calculated heat release rate curve was used as an indata-file in the program Fire Dynamics Simulator. In the program two simulations were conducted for two different scenarios. The first simulation shows how long the fire will continue with available oxygen. The second simulation shows a scenario with broken windows leading to the fire becoming ventilation controlled. Results from the simulations were compared with the calculations.Assumptions were made on the fuel content of the wires. The amount of polyethylene was assumed to be two-thirds and the rest polyvinylchloride. To assess the effect of these assumptions calculations were made with cables consisting fully of each fuel content. This sensitivity analysis showed that there became differences in the results but the consequences for the area remained the same. Adoption of time until the fire reaches its maximum was based on tests conducted in a large study of the Swedish Technical Research institution (SP). This study showed that the given test setup, horizontal scenario, reaches its maximum effect after about 15 minutes. In the sensitivity analysis a lower time than this was used, which lead to a higher maximum power. But this change is not expected to affect the consequences of the fire.The discussion dealt with why the analyse-term estimated heat release rate curve did not fit with the simulated in FDS. The discussion also provides a list for improvement.• A complete cleanup of the culvert and removal of remains.• Supervision process for the culvert needs to be established. There could be a quick overview in conjunction with other inspections, to continuously maintain a clean environment in the culvert.• Repair of the broken lights.• Inventory and removal of non-utilized cables in the culvert.• Cable and pipe section between B and C should be reviewed to ensure that the fire spread not will reach the container hall (see BBR 5:56)• Cover plates over cable ladder with high voltage cables can be installed to reduce the possibility of fire, to protect the cables if a fire still occurs and to delay the ignition time.• A contingency plan for the culvert should be established. This should include a drawing over the culvert.• Water and sewage pipes should be located outside of the culvert. Even old unused sewage pipes should be removed because they may contain fluid, which could lead to short circuiting of high voltage cables if it starts to leak.• Some form of detection in the transformer housing could be justified. From here, the whole area is supplied with electricity and urgency of protecting this should be high on a priority list. As it is today, there is no detection.

Place, publisher, year, edition, pages
2013. , 53 p.
Keyword [en]
Keyword [sv]
Teknik, Kabelkulvert
URN: urn:nbn:se:ltu:diva-58186Local ID: ec5a2555-2217-47f2-ac87-7faf7ca33ae2OAI: diva2:1031574
Subject / course
Student thesis, at least 15 credits
Educational program
Fire Protection Engineer, bachelor's level
Validerat; 20130104 (global_studentproject_submitter)Available from: 2016-10-04 Created: 2016-10-04Bibliographically approved

Open Access in DiVA

fulltext(12730 kB)0 downloads
File information
File name FULLTEXT02.pdfFile size 12730 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Search in DiVA

By author/editor
Hörnqvist, Anton

Search outside of DiVA

GoogleGoogle Scholar
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

ReferencesLink to record
Permanent link

Direct link