Madeleine – Sky special – 2 May 2017

On 2 May 2017, Sky showed ‘Searching for Madeleine’, a special to mark the10th anniversary of the disappearance of Madeleine McCann. The presenter was Martin Brunt, who has followed the case for the 10 years since it began. The studio guest was Colin Sutton, an ex-DCI from Scotland Yard with experience of conducting major investigations.

The fist 10 minutes covered the basics. The holiday, the Tapas zone, the initial response to the incident by Portuguese police.

Sky News on 4 May 2007 ran with the story that a 3 year old British girl was missing on the Algarve. Pedro do Carmo, Deputy Director, Judicial Police, described the initial work as a rescue operation, looking for a child that was missing.

Here Sky hit its first wobbly. It says the apartment was let out twice before it was sealed off for a full forensic examination. The reality is different. The PJ from Portimão tried to collect forensic evidence in the very early hours of 4 May 2007. Irene Trovão, also a local forensic officer, was videoed checking the shutter of the children’s bedroom for fingerprints. And while Gerry and Kate McCann were giving their first witness statements, a forensics duo from Lisbon conducted the major forensic examination on the afternoon of 4 May 2007. The forensics had been done. There was no way to foresee the apartment should be sealed off until Eddie and Keela were deployed.

The centrepiece of the Sky programme was a Home Office report written by Jim Gamble, then head of CEOP, the Child Exploitation and Online Protection Centre.

This documented the many organisations that were involved close to the beginning, and the difficulties this caused. Alan Johnson, the Home Secretary at the time, questioned if Leicestershire Police had the wherewithal to handle this type of investigation. Mr Gamble was asked to consider if it was worth getting Scotland Yard involved. Mr Gamble suggested a scoping review to identify if opportunities had been missed, but officials appeared to be set against this.

Mr Gamble was shocked to find the parents had not been investigated first by the Portuguese police, in order to clear the ground for further enquiries. He went on to say the Portuguese response was inadequate, but he used a comparison in the UK that does not approximate to the situation in Luz in 2007. I will return to that in a future post.

Colin Sutton made the point that a snapshot of the incident area was not constructed, and more could have been done by UK police re interviewing British holidaymakers who had returned to the UK, and British workers in the ‘complex’.

My main criticism of the early effort is that apparently little was done to get door-to-door information in the immediate vicinity of apartment 5A.

Sky went on to cover leaks to the Portuguese press, concerning dog alerts and supposed DNA results. Mr Sutton pointed out that dog alerts are not evidence.

The events around the McCanns being made arguidos, flying home to the UK, and removal of arguido status upon archiving of the case was covered.

There appeared to be a 3-way split between the McCanns, the Portuguese police and the UK police. The CEOP report then makes an odd assertion. It alleges the McCanns had a significant amount of information from their private investigators, and this information had not been fully shared with either the Portuguese police or the UK police. I cannot see how Mr Gamble could reach such a conclusion. Perhaps it is explained in the CEOP report, but I haven’t read that document.

Mark Rowley, Assistant Commissioner, Metropolitan Police, explained there had been a discussion of the case in 2011 between the Prime Ministers of Portugal and the UK, and it was agreed that Scotland Yard would get involved.

The documentary then covered the remit. Colin Sutton explained that a fresh investigation should start right at the beginning. This echoes what was said by Jim Gamble. However, Operation Grange was to be restricted to abduction. AC Mark Rowley says parental involvement had been covered by the original Portuguese investigation. The recent Supreme Court decision made it clear this is not the case.

The Sky documentary moved on to the Jane Tanner sighting. Martin Brunt pointed out the obvious – namely if the man was coming from the Ocean Club night crèche, then he was going the wrong way. Jane Tanner’s rogatory statement pointed out this problem. If the night crèche closed at 11.30pm, It is actually more likely that at 9.15pm, the time of the Tanner sighting, he was heading towards the night crèche.

Scotland Yard presented two e-fits of a man carrying a child ‘towards the beach’. This of course was the Smith sighting at 10pm. Crimewatch 2013 did indeed state this man was heading towards the beach.

This suggests that Martin Brunt does not fully understand the Smith sighting. 12-year-old Aoife Smith’s statement does not fit with ‘towards the beach’. Should Mr Brunt ever return to Luz, I will be happy to show him why Aoife Smith’s statement strongly suggests ‘towards the beach’ is wrong. And why that man is likely to be Portuguese and innocent. Plus why that man is unlikely to come forward. And what needs to be done to get him to identify himself.

The documentary covered Operation Grange’s look at charity collectors. There is an easy test for this. The bogus ones do door-to-door, and disappear rapidly if they make some cash. The genuine ones go to the main thoroughfares and work there for hours on end.

Then Sky covered a burglary gone wrong. Whilst Operation Grange evaluated this as viable, Portuguese police did not think it likely.

The documentary moved to mobile phone data. The CEOP report says there was lots of it, but it was badly handled by Portuguese investigators. It had not been fully analysed, and the Portuguese should accept UK help. This sounds to me to be very over-simplistic, but I cannot be certain as I have not read the CEOP report.

Then the documentary moved to its weakest point, what can be extracted from that phone data. Nothing Colin Sutton said on this has much relevance to Luz on 3 May 2007.

As is normal, there were 3 cellphone operators in Luz – Optimus, TMN and Vodafone. Roughly speaking, each operator cuts Luz into a western half and an eastern half, and that is as much as you get. Was the cellphone active in Luz that night, and if so, was it in the west of Luz or the east.

Take for example Kate McCann. Her phone was active that night on Optimus antenna Luz 2. That antenna covers the east of Luz, and apartment 5A is indeed in the east of Luz. But the whole of the Ocean Club is in the eastern half of Luz, as is the majority of the commercial establishments e.g. the Mirage. I cannot tell from phone data if Kate was in or around 5A when her phone was active. The phone data is very rough.

Further, DCI Andy Redwood has said that a major obstacle to phone data analysis was PAYG phones.

4 people were made arguidos in July 2014, but have now been informed they are no longer persons of interest.

The new Portuguese investigation focussed on a series of sex attacks in the Algarve. It would appear most were on older children, but one was on a child aged 3. Euclides Monteiro, an ex-waiter at the Ocean Club, was identified by the Portuguese investigation as a suspect for the sex attacks. DNA tests ruled out Mr Monteiro. He had been killed in a tractor accident in 2009.

The Sky documentary examined the woke and wandered theory. Local ex-pat Mr John Ballinger provided some photos of the road works in Luz around that time. There was no examination as to why Kate McCann’s description of apartment 5A that night is a poor fit with woke and wandered.

Mr Brunt pointed out that there is no evidence to prove Madeleine came to any harm, so she may still be alive.

Have lessons been learned from the disappearance of Madeleine McCann? Jim Gamble and Alan Johnson think not.

The documentary covered some of the Internet abuse directed at Kate and Gerry. Two police investigations found no evidence of their involvement in Madeleine’s disappearance. The Sky investigation also found no such evidence.

It concluded that the mystery of what happened to Madeleine McCann remains just that. A mystery.

AC Mark Rowley said there is a significant line of enquiry that remains to be pursued, but would not divulge what it was.

On the armchair experts forum that I prefer, the general view was that little was learned from this Sky special. However, that is not the correct view to take, in my opinion. This programme was not aimed at a handful of amateur detectives. It was targeting the greater British public. And for those, I suspect the key point that was delivered was that roughly £12 million down the line, the investigation is fatally flawed because, despite what DCI Andy Redwood said, it did not start by going back to the very beginning.


Madeleine – 3 phone operators

The 3 phone operators in Luz in May 2007 were Optimus, TMN and Vodafone.

The Tapas 9 phone operators were as follows

David Payne 7748 Vodafone UK

Dianne Webster 7790 Orange

Fiona Payne no mobile 2-4 May

Jane Tanner 7808 Telefonica UK

Gerry McCann 7786 Vodafone UK

Kate McCann 7903 EE ltd (TM)

Matthew Oldfield 7771 Vodafone UK

Rachael Oldfield 7771 Vodafone UK

Russell O’Brien 7713 Telefonica UK

So far so good, but this is where the going gets tough. Trying to match up Optimus and TMN with Orange, Telefonica UK, and EE ltd is proving puzzling.

Telefonica is Spanish, and does not in 2015 trade in Portugal, so there must have been a deal in place to run through one of the 3 operators in Portugal, but which one or ones?

At this point in time, we can reasonably assume that a Vodapfone UK mobile would go through the Vodafone, so at least we start with 4 of the 8 people.

We also know that when the caller and the person called are both within Luz and the call goes through two operators, we see double reporting of a single call, once each by the two operators.

Gerry makes 2 calls to Kate on 3 May 2007, and both of these turn up twice in the records. So Kate is not on the same operator as Gerry i.e. she is not going through Vodafone. This is confirmed on 4 May 2007 when David Payne, on Vodafone, is in contact with Kate, not on Vodafone, and the call gets double reported.

Inspector Paulo Dias says that Optimus contacts are marked in orange (the colour) while TMN contacts are in blue. I cannot find anything that states what colour is used for Vodafone.

I also have the problem that I don’t see how one can tell which network a 3rd party is using. I can tell that Gerry is on Vodafone, but when he connects to 2 English numbers he has in common with Kate, which colour is appropriate? Should it be the Vodafone colour that he is on? Should it be Kate’s orange or blue (even assuming I can work out which she is on?)

For the moment, here is a slightly updated graphic of the shared 3rd parties. To match the apparent heaviness of the lines, I have used orange, a lightish blue, and a dark green for Vodafone.

Shared numbers 2

Madeleine – 74,000 calls/texts – Kate part 2

Page 23

Kate after the dialogue with the user of that number, still on that day 2, exchanges messages with the number 4479xxxx7624, by 20:08:30, 20:08:41 she receives. On day 3, she sends a message by 08:23:22 hours, receiving the response to that contact by 08:24:33 hours. Following this exchange of messages, she returns to do it with the number 4479xxxx6678, already treated above.

With the disappearance of Madeleine there is a strong increase in her traffic: from 12 contacts that she held until it passed to 132 carried out in the period understood between 23:14:51 hours (contact with her husband), up to 23:57:57 hours, on day 4.

Also she makes contact with the phone 4411xxxx3138, at 06:05:29 hours, day 4, maintaining a 702 second dialogue. This number does not activate any antenna, however, it relates to the number 4479xxxx6678, to which reference has already been made above as being the common contact between Russell and Kate

She makes a call to the number 4479xxxx1027, at 23:37:01 hours, of day 4 (in the graphic there arise two calls with the interval of one second, referring to the same call, agreeing with the two records, made in the calls of Kate and recorded in the number in question).

The number 4479xxxx1027, contacts with the numbers 964078520 and 447812128964. This latter activates an antenna TMN for the first time by 19:35:33 hours, of day 4. It is unknown to whom it may belong. The first makes its appearance when it receives a phone call from mobile phone 962xxx835, on day 4, at 19:50:19 hours.

On the 4th, at 00:47:23 hours, she receives a call from phone 289xxx562, with a duration of 239 seconds. By 01:02:08, that same day, she turns to receive a new call from the same number, maintaining a dialogue of 223 seconds. {That’s the Pacífico Bar in Vilamoura.}

In Table 11 are records of the communications of Kate Healy. {Unfortunately, the table is sorted in caller number order, making it hard to work out what is going on.}

Madeleine – 74,000 calls/texts – Rachael Oldfield

Page 14

Rachael Manpilly has a contact in common with Jane Tanner, the phone 917453319, that which its first registration is 22:06:52 hours of the day 4. This national number also contacts with 441274585431. This does not trigger any antenna studied.

The only member of the group with whom she maintains contacts is with her husband.

The first record contained in the database is a message it receives from number 1100, at 09:57:19 hours, of day 2. Subsequently, it exchanges 7 messages (6 with the number 447812799092 and 1 with the number 447817049612; these figures do not activate any antenna of those studied, nor have other contacts to do so). One of these, the last before the alert being given, occurred at 16:06:51 hours, of day 3.

On day 2, at 09:57:32 hours, she makes a call to telephone 447973100123 of which lasts 101 seconds (00:01:40 hours). This number does not trigger any of the analysed antenna serving the studied area, however it relates to the following numbers broken down, which activate antennas of TMN and Optimus: 447711219039 (1st activation: 08:11:43 hours, day 2) , 447778646683 (1st activation: 09:49:13 hours, day 2) …

{There follows a long list of 1st activations, all on 2 May, except for the last on 3 May.

See }

In the graphic of this element of the group, there is also her momentary absence to which reference has already been made, when the graphic of Matthew was reported.

The other contacts are made with numbers which are not influenced by any of the antennas. There are also links to the message centres of the operators.

The first contact that she effects after the alert, after exchanging contact with her husband, is a call that she makes to the number 447768698713, at 01:42:55 hours, with a duration of 75 seconds. This number returns the contact, at 01:58:53 hours with a duration of 6 minutes and 8 seconds. As already mentioned this number does not activate any antenna in that zone of the country.

The call receiving or perform with greater time is recorded on day 4, at 2:33:38 pm, from 44549370165752 number, which lasts 692 seconds.

In Table 7 are the records of the Rachael Manpilly communications.

{Table 7 is at

This is a big table and has some unusual looking numbers in it, so it is going to take a bit of working through.

The comment on Matthew´s records that Rachael was outside of Luz or had her phone switched off at 00:08:01 on 4 May is still puzzling. Rachael’s records seem to show that she was in Luz and her phone was switched on, so more checking is required here.}

Madeleine – 74,000 calls/texts – Matthew Oldfield

Page 13

Matthew Oldfield makes no contact before the alert is given. The first record of the three days is an SMS that he receives, from the number TMobileej (this value is unknown, however it is in the data base sent by TMN) for 22:45:05 hours of day 3.

The person with whom he contacts more is with his wife Rachael Manpilly, who, by 00:08:01 hours, of day 4 would be outside the radius of action of the antennas of Praia da Luz, or be with the phone off. This contact between the two appears only in Matthew’s records.

One of the situations that have any interest to emphasize is that he receives an SMS number 447958879879, by 00:08:12, day 4, which according to Vodaphone databases was in the UK. This number does not trigger the antennas of Praia da Luz, but contacts with numbers 447773335588, 447930535254, 44797018745, 447984763199, 447985111885, 447985609362, 447986976122, that yes activate antennas of Praia da Luz, some even on May 2, as is the case of number 447986976122 (first constant activation of the TMN database is on 02-05-2007, at 09:14:42 hours); and the number 447985111885 (first activation on 02-05-2007, at 13:04:46 hours). The other five only activate antennas of Praia da Luz on the day 4. These phones shall not contact each other.

In table 6 are registered the communications of Matthew Oldfield.

{Table 6 is at

There are two pairs of overlapping phone calls on 3 May 2007at 22:51 and 23:06.

The comment about Rachael being out of Luz or having her mobile switched off seems odd. Perhaps there’s more in her phone analysis.}

Madeleine v Luz phone traffic

The PJ obtained files of phone traffic for Luz from all 3 operators (Optimus, TMN, Vodaphone) for the dates 2 May 2007 to 4 May 2007. This amounts to over 74,000 calls/texts in total, according to the PJ Files.

On 4 Dec 2013, Danny Shaw wrote an article for the BBC covering the status of Operation Grange, and headlined it “Madeleine McCann: Phone records may hold key, UK police say”.

The article includes lines of enquiry other than the Luz phone traffic, so here is what was said purely about that phone data.

Mobile phone records may hold the key to solving the disappearance of Madeleine McCann, detectives believe.

Scotland Yard officers are analysing data from phones belonging to people in the village at the time – 41 people of interest have been identified.

Detective Chief Inspector Andy Redwood, who is leading the inquiry, said officers were examining a “substantial amount of data” from thousands of mobile phones thought to belong to people who were in the resort of Praia da Luz in the days just before, during and after Madeleine’s disappearance.

Police are trying to identify the owner of each phone to build up a picture of exactly who was in the area. More than 3,000 people live in Praia da Luz, while holidaymakers and seasonal workers visit from countries across the world.

“This is not just a general trawl,” said Det Ch Insp Redwood.

“It’s a targeted attack on that data to see if it assists us to find out what happened to Madeleine McCann at that time.”

‘Call timeline’

Det Ch Insp Redwood said officers had so far been unable to attribute a “large number” of mobile numbers and admitted that it was difficult to do so with phones bought six years ago on a pay-as-you-go basis.

The records also contain information on which phone numbers were dialled and when calls were made. It is thought some phone numbers might appear on police intelligence systems or be linked to criminals.

“We can see what the phone is doing, but we can’t see the text messages,” said the detective. “It shows a timeline of the call data.”

According to Scotland Yard, the phone records had been “looked at” during the initial Portuguese police investigation but not in detail.

Asked by reporters if the information held the key to the investigation, Det Ch Insp Redwood replied “It could do.”

Speaking to the BBC’s Today programme, Peter Sommer, an expert on cyber security, said the “multi-jurisdictional nature” of the case, which would involve mobile phone companies in different countries – and the gap in time – could make it harder to track people down.

But he said “cell site data” was routinely used in most criminal court cases in the UK.

Mr Gamble said the EU data retention directive, which compels telephone companies to retain call and internet records for a period of time, was at an “immature stage” in 2007.

But he said it appeared the data “wasn’t properly or appropriately interrogated,” at the time.

In UK investigations, he would expect the data to have been examined almost immediately, he said, but the “complex nature and geography” had made it more difficult.

There is a sidebar on how this data can be interpreted. What can phone data tell us?

By Matthew Wall Technology reporter, BBC News

As mobile phones constantly send and receive data from mobile phone masts, a user’s location can be identified to within a few hundred metres using triangulation techniques. Modern smartphones with GPS built in can be located far more accurately.

Mobile phone records include the numbers of the call sender and receiver, the call duration and time. Couple this with location information and you can establish where and when callers made or received a call. This information is often used to verify or knock down alibis in criminal cases.

The difficulty for investigators is establishing the identity of the user if the phone is pay-as-you-go (PAYG) rather than on a pay-monthly contract linked to a bank account. PAYG phones, SIM cards and top-up cards can be bought in-store for cash, leaving no identifying trail for investigators to follow.

And because such phones can be lent or sold to other people, establishing exactly who made a telephone call is made even more difficult.

This is where the BBC report ends.

Let me deal with the part by Matthew Wall first. I assume he is not familiar with the PJ files, otherwise he would be aware of the analyses of data carried out with respect to the Tapas 9. None of these had information that could be used to triangulate. The closest to triangulation was when a signal passed from one mast to a second, and a reasonable assumption could be made in some instances about the location of the phone.

This triangulation idea was touted back in December 2007, when it was learned that a vast volume of phone data had been collected by the PJ team. Hailed by the media then as a breakthrough, a casual reader should deduce that it was no such thing then, and that Operation Grange has encountered difficulties in processing the data since.

The PJ files have data that appears to cover – call sender number, call receiver number, date and time, mast used, call duration, and which antenna of the mast was activated.

The antenna that was activated gives a very rough direction relating the mast to the mobile phone. A knowledge of other masts in the region allows this to be narrowed to an area, but that area is normally large. Two masts for the same operator can cover a distance of 5km (built-up) to 8km (rural), so the swathe covered by the active mast antenna is not small.

I am not aware of the procedures used in various countries at the time re PAYG phones, but the major countries involved are Portugal, Spain, the United Kingdom, Ireland, Germany and Holland. Then there is a host of minor countries to consider.

Let me turn now to checks re phone numbers of known criminals. If that angle has produced interesting lines of enquiry then it has taken a long time for Operation Grange to act on these.

Mr Gamble’s comments seem irrelevant. The three operators, Optimus, TMN and Vodaphone, provided all that one would expect to get in such a situation. Then applying a theoretical UK approach to a multi-country problem in Portugal involving this number of phone messages, when a comparable situation does not appear to have ever happened in the UK, does not enlighten the reader.

The problem is that Operation Grange is trying to tackle 74,000 calls and texts in an out of Luz over the 3 days of 2 May 2007 to 4 May 2007.

The bulk of data before the Madeleine story was widely aired in the media can reasonably be assumed to be driven in the main by things that do NOT relate to Madeleine’s disappearance. If there IS traffic relating to her disappearance before the news broke, it will be masked by what was going on in Luz in the period of 2 and 3 May.

Here is where Operation Grange hits a major problem. To make sense of this, to get the ability to strip off the normal traffic and identify the abnormal traffic, the OG analyst has to understand the normal workings of Luz.

Let me illustrate this with a trivial example. When we get a take-away from the Luz Chinese restaurant, we phone the order through, ask how long it will take, and pick it up at the appropriate time.

This is not a major breakthrough. It would cover a small number of phone calls made for take-aways and a few more for restaurant reservations.

It simply illustrates the point that to understand the traffic, one has to understand Luz.

Apart from a visit to dig up central Luz in mid-2014, Operation Grange does not have that experience of understanding.

Then a second obstacle is encountered, namely, that European legislation prevents the distribution of this type of phone traffic information to private individuals.

This means that people with a knowledge of Luz, such as myself, will never be allowed to see the data. Since Portuguese law prevents private individuals from investigating crimes in Portugal, my potential input is severely limited.

If Operation Grange is shelved, and the McCanns appoint more private investigators, they will be in a position to investigate from outside Portugal i.e. one of the blocks is removed. However, the other blocks are not. The McCanns and their private investigators will not be allowed access to this phone traffic data, and the private investigators will lack knowledge of the workings of Luz. That would mean the people with the legal ability to progress this angle, Scotland Yard, are at the stage of not doing so, whilst those who are active on the case are denied access to valuable information.

A bit of lateral thinking suggests the following. If Operation Grange won’t come to Luz, then Luz must go to Operation Grange.

PS On 20 Dec 07, the Mail Online reported that PJ officers were checking hundreds of residents in Luz re calls made on 3 May 2007, from 9:30pm onwards. This is the report with the triangulation story, which is erroneous. The geographical scope of masts covered also looks suspect. The PJ would be able, one presumes, to locate those people on a phone contract registered in Portugal. As to the rest…

Madeleine – Luz pie

How does the Luz pie cut up on the telephone data?

There were 3 mobile operators in Luz in May 2007 – Optimus, TMN and Vodaphone. Other companies may have been working as virtual operators – those who supplied a service but did it by simply renting bandwidth from one of the three key players.

So, I want to focus on the big 3 and what is available from the information on the table.

Heriberto has detailed the latitude and longitude of the masts in Luz, and the direction of each main beam from the antennae.

I have cut up Luz based on a simple assumption. If a mast has 2 lobes, one pointing west and one pointing east, then there will be a boundary between these two. That border will split the zones covered by the lobes.

On the following graphic, apartment 5A is the pink dot. The lines are the boundaries between lobes for each mast.

Luz pie

The green lines are the areas covered by the Optimus mast, located on the Luz water tower. It has 3 (unequal) segments. It is poor for analysing movement within Luz as one lobe covers the bulk of the town.

The yellow lines are the TMN mast. It is on the top of 2 Rua Primeiro de Maio. This has 2 lobes, pointing east-west, so the boundary is basically north-south. The boundary is somewhere around block 5, but the graphic is not meant to be 100% accurate. It gives the basic idea, but there must be a wobble factor.

The red lines are the Vodaphone mast boundaries. Heriberto puts the mast slightly elsewhere, but I think it is also on top of 2 Rua Primeiro de Maio. It is 2 part, and slightly angled to the TMN mast.

It cuts the Luz pie into block 5 plus Tapas 9 to the east, with the Dolphin, Kelly’s and Smithman to the west.

I don’t know what is in the Scotland Yard data, so I don’t know how detailed they can delve.

But I wonder if this gives me enough to hunt down the Scandal at Chaplins? Is this a game changer?