Hello Charlotte,
1. Data management issue
From your description below it is not clear what data management steps you have used to track separated parents. So, I would like to point out the following points about the survey and the data which may help you in checking if what you have done is correct:
(a) For any person in the sample, the value of the mpid (fpid) will be =-8 in the wave when they are not living in the same household as their mother (father) and will be greater than zero (a positive valid PIDP) in the wave they are living in the same household with them. So, for example, if child A was living with both her father and mother in Wave 1 but in Wave 2 she was living with her mother only, then for her mpid > -8 in both Waves 1 and 2, fpid>-8 in Wave 1 and fpid=-8 in Wave 2.
(b) You are correct in assuming that if the child has the same w_hidp as one of the parents but not the other parent then that means the parents are not living together. In most cases this would mean the parents are separated but in a few cases this may be because one of the parent has moved somewhere for work reasons depending on how the respondent describes the situation. For example, if someone works away part of the week then they will most likely be counted as co-resident. On the other hand, if someone is away for a much longer duration such that they will not be at this residence during the entire fieldwork period then they will be considered as a split off household and attempts will be made to interview them at that other address. So, to make sure, please check the marital status (w_mastat_dv) of the co-resident parent. If they are really separated then this should show up as separated or divorced (and the equivalent codes for those who were in civil partnerships).
I found that of the 13123 sample members who were living with both parents in Wave 1 and were also enumerated in Wave 2, 96% were still living with both parents in Wave 2, 2% with only mother, 0.3% with only father and 1.5% with neither.
2. Tracking of separated partners and surveying their new household members. To answer this I will have to describe the following rules. See Section 2.2.6 in the User Guide for a complete description. Here is a summary:
Everyone in the Wave 1 sample households (with some exceptions for the non-ethnic minority members in the boost samples, EMBS & IEMBS) are considered to be Original Sample Members (OSMs). In every subsequent wave attempts are made to follow every OSM (children of OSM mothers are also OSMs) wherever they move as long as it is within the UK. From Wave 2 onwards, any person moving into a household with at least one OSM is also interviewed. These new household members are called Temporary Sample Members (TSMs) and they are not followed if they move out of a household with OSMs. But if a TSM father has a child with an OSM mother then the TSM father’s status changes to a Permanent Sample Member. PSMs have the same following rules as an OSM, that is, they are followed wherever they move as long as it is within UK.
In your example, if the separated parent is an OSM or PSM then their new partners and their children will be surveyed as long as they are living in the household with this separated OSM/PSM parent.
Best wishes,
Alita
Updated by Stephanie Auty 
Updated by